At the end of June, Apple released the iOS 26.5.2 update, containing a long list of security fixes. The timing of the update differed from Apple’s typical approach, which sees the iPhone maker issue mid-cycle patches for emergency security fixes or urgent bugs only.

Yet when iOS 26.5.2 was released, Apple told Reuters the move was part of an overhaul of its strategy as AI tools find vulnerabilities at scale. Indeed, some of the flaws patched in its update were found using AI, painting a clear picture of what security teams can expect in the future.

Apple is not the only vendor to supercharge its patch release cycle fuelled by AI-enabled vulnerability discovery. This June was Microsoft’s biggest Patch Tuesday ever, fixing over 200 bugs, while a recent Google Chrome patch fixed 400 issues in the browser.

As ITPro has already reported, patching velocity must increase as Anthropic’s frontier AI model Claude Mythos and others like it start to find flaws en masse.

But Apple’s accelerated release schedule highlights the impact on mobile devices – especially when part of a bring your own device (BYOD) policy. What can security leaders do to regain control?

Shrinking attack windows

Experts say shrinking attack windows are making patch management an urgent problem. The time between a flaw being discovered and weaponized is “falling dramatically” as AI tools “accelerate both research and exploitation”, says Tristan Shortland, CTO at Infinity Group.

“Traditional patch cycles that measured response times in weeks are increasingly at odds with a threat environment that moves in hours.”

With this in mind, most industry commentators see the move towards accelerated patch releases as positive. “From a risk perspective, I'd much rather see organizations patch quickly and in a controlled way than wait for monthly cycles and leave vulnerabilities sitting there, waiting to be exploited,” says Nathan Davies-Webb, principal consultant at Acumen Cyber.

Yet rapid patch release cycles add pressure in scenarios where employees use personal devices to access corporate networks. “Apple, Google and Microsoft can remind users to update, but they don’t always enforce it,” points out Jake Moore, global cybersecurity advisor at ESET.

For those that run BYOD policies, the impact of late patching could be vast, agrees Sriram Kakalara, chief product officer at Scalefusion. “Personal iPhones and iPads often access corporate apps and data from outside the IT team's direct control,” he warns.

More updates, less warning

Another challenge is the sheer number of mobile fleets to be updated. The trend towards increasing numbers of mobile device patches could be a problem for larger distributed organizations such as retailers, according to Kakalara.

“These might have thousands of devices running on iOS, and every accelerated release means pushing updates more often and with less warning,” he says.

He points out that many organizations “fail to scale their patching and access strategies as their endpoint estate expands, leaving known vulnerabilities unresolved”.

The onus is now on organizations to operate “faster and more efficiently”, he says.

Beyond simply deploying updates more rapidly, the challenge will be ensuring every endpoint has received them, adds Andy Ward, SVP International at Absolute Security.

“The rise of hybrid working over the last decade has made this increasingly difficult. Remote work and BYOD leave organizations with gaps in visibility that attackers can easily exploit. Every day a critical patch is delayed extends the window of opportunity for cybercriminals.”

While there’s no doubt AI can be beneficial in vulnerability discovery, the pressure on patching as a result of the constant flurry of security fixes is only going to get worse.

Shortland expects more vendors will follow Apple's lead, issuing patches more rapidly in line with vulnerability discovery. “AI is exposing vulnerabilities at a scale we've not seen before, which means the industry will have to move towards more continuous patching and risk-based remediation models, rather than relying on predictable release cadences.”

Security teams should treat Apple's move as “the new normal”, agrees Kakalara. “The precedent means patches can now arrive unexpectedly and without notice, so waiting for a scheduled maintenance window is no longer a viable strategy. Teams need to be on permanent alert, ready to test and deploy at speed.”

No more patch day

It might seem complex, but there are some simple steps security leaders can take to ensure fixes are applied as soon as they’re released, including to BYOD devices.

Rather than banning BYOD, Moore thinks the solution is implementing conditional access controls to ensure outdated devices can’t access company systems. “This means it becomes nearly impossible to have devices running outdated software.”

But at the same time, companies need to avoid treating every new flaw as an emergency. Instead, the focus should be on “what is actually being exploited – rather than what is possible”, Moore advises.

Ian Thornton-Trump, CISO at Inversion6, believes security leaders must make themselves aware of the size of their company’s attack surface, and be able to identify legacy devices “either properly supported by vendors, or unsupported”.

In tandem, Thornton-Trump outlines the importance of firms testing their own defences. “Prioritize red teaming and penetration testing to try and find weaknesses before bad guys do, and before security researchers encourage firms to respond.”

Moore recommends automated patching as an ideal scenario. “Long gone are the days where Tuesday was the patch day,” he says. “Manual processes were built for a far slower time, but as we shift into the AI era, we are going to have to adapt and evolve as quickly as the technology does.”

Davies-Webb concurs. He thinks the point has been reached where not using automatic patching “is becoming harder to justify”.

“There was a time when concerns around stability and functionality made organizations cautious about patching too quickly, but for most, the risk now sits firmly on the other side of the equation,” says Davies-Webb.

“AI is accelerating how quickly vulnerabilities are discovered and weaponized, which means organizations need to be reducing exposure as quickly as they can."

PC shipments have fallen for the first time in more than two years, hit by the memory supply crunch caused in part by the rise of AI.

Recent analysis from IDC found global PC shipments slid 4.9% on year in the second quarter of this year to 68.2 million units, down from 71.7m in the same quarter last year.

The figures mark the first decline in the market after nine quarters of growth, and that’s largely down to continued memory shortages.

IDC noted, however, that the situation has been exacerbated by supply issues with storage and other components, as well as what the company referred to as "geopolitical issues".

The massive AI infrastructure build-out has led to a shortage in memory, spurred on by manufacturers pivoting to produce more lucrative data center components rather than those for consumer products.

That RAM shortage has led to a sharp decline in lower cost devices, with Gartner predicting earlier this year that PC sales will drop by 10%.

Consumers picking up the bill

IDC said that while units shipped slid, PC makers reported growth in revenue, showing how the cost of such challenges are being passed on to consumers and businesses.

"The real story here is the disconnect between units and dollars: shipments are falling, but revenue is climbing because vendors are pushing through price increases faster than demand is dropping," said Jitesh Ubrani, research director for consumer devices at IDC.

That's not likely to change anytime soon, Ubrani warned.

"Given worsening macro conditions and a memory shortage that isn’t expected to ease until early 2028, we don’t expect another round of inventory pull-forward, which points to a sharp slowdown in growth rates in the second half of 2026," he added.

"Vendors are bracing for further price hikes into 2027, and channels are already flagging concern about elevated inventory at these higher price points."

A host of device manufacturers have raised prices in response to the spiralling memory crisis. Outgoing Apple CEO Tim Cook said last month that the company would raise prices due to the issue.

"There's less supply at a time when consumers want devices and the memory guys are passing along huge price increases," he said at the time.

Apple wins with MacBook Neo

The increased pressure on brands could lead to vendor consolidation of market share, IDC warned. The consultancy noted that larger brands such as Apple, Dell, or Lenovo can use their scale across different product lines – be it smartphones or servers – to access memory supply.

The result here is that smaller rivals could be left in the dust and scrambling for what supplies are left.

"As market conditions continue to worsen… supply chain management and capabilities are increasingly important," said Jean Philippe Bouchard, vice president for consumer devices at IDC.

"The largest vendors, with their buying power and long-standing supplier ties, are best positioned to take share from smaller rivals."

Compared to the second quarter in 2025, the big three PC makers – Lenovo, HP, and Dell – were largely flat in terms of market share, though all three saw single-digit decreases in shipments, with HP in particular down 9%.

Apple, however, grew market share from 8.5% to 9.9%, with shipments up 10% from 6.1m to 6.7m in the second quarter of this year versus 2025.

"Apple’s share gain coincided with its latest product launch, the MacBook Neo, and while the company did raise prices in line with the broader market, it still remains well positioned against rivals facing the same cost pressures," added Bouchard.

FOLLOW US ON SOCIAL MEDIA

Apple has dropped an unexpected out-of-band patch, and it's apparently because of concerns that AI is helping hackers exploit vulnerabilities at lightning speed.

The tech giant released iOS 25.6.2 yesterday, with fixes for iOS and iPadOS, despite normally bundling such security patches into wider updates. Apple hasn't said when its next major update, iOS 26.6, is set to be released, but it's widely expected in the next few weeks ahead of iOS 27 in the autumn.

Apple told Reuters that the extra security update was down to concerns about AI shortening the time between a flaw being spotted and exploited by hackers.

The company said that shorter timelines from flaw to exploit meant it needed to respond by cutting the time between when updates are announced and released.

That shouldn't be taken as a suggestion that Apple was seeing any evidence that hackers were already making use of any of the flaws included in the update, the company noted.

ITPro approached Apple for confirmation of the changes made to its security patching policy, but did not receive a response by time of publication.

Apple responds to accelerating threats

The move comes in response to wider concerns that AI is helping hackers find flaws more easily and turn them into exploits more quickly. Indeed, this has been a recurring talking point since the release of powerful new frontier models such as Anthropic’s Claude Mythos range.

Last month, Google said it had spotted cyber criminals using AI to build a working zero-day exploit, with John Hultquist, chief analyst at GTIG, warning that an “AI vulnerability race is imminent”

That echoes a warning earlier this year from Daniel dos Santos, VP of research at Forescout, who told ITPro that enterprises should brace themselves for an explosion of vulnerabilities thanks to AI, further adding to the workload of security teams.

Jake Moore, Global Cybersecurity Advisor at ESET, told ITPro that while the “jury is still out” on how powerful tools such as Mythos are, AI advances do mean that security teams and threat actors alike are scrambling to find software vulnerabilities.

"Whenever a new issue is discovered, there's a race to patch flaws before they can be exploited – and if users or security teams hold off on installing updates, it's a gift to threat actors,” he said.

Moore added that increasing numbers of patches mean the process needs to be automated as much as possible to avoid update fatigue for both end users and security practitioners.

"In the age of automated vulnerability discovery, we're past the point of expecting users and security teams in organisations to manually patch everything," he commented.

"Updates should be automated wherever possible to reduce the burden on users while making sure patches are applied rapidly."

Apple iOS 26.5.2

The security fixes released by Apple were previously revealed via a beta of the update, as is Apple's usual procedure, and addressed more than 30 vulnerabilities across iOS and iPadOS.

Those include kernel vulnerabilities that could corrupt memory and trigger system shutdown and a stack of WebKit issues, including one that could leak sensitive data after visiting a website.

Three WebKit flaws were spotted by OpenAI's Codex Security, while another was spotted by a pair of AI researchers using Claude Anthropic.

The update should have already landed on devices, and is supported on these devices:

• iPhone 11 and later
• iPad Pro 12.9 inch, 3rd generation and later
• iPad Pro 11 inch, 1st generation and later
• iPad Air 3rd generation and later
• iPad 8th generation and later
• iPad mini 5th generation and later

FOLLOW US ON SOCIAL MEDIA

Anyone waiting for September's annual iPhone launch will be unhappy with this news: Apple CEO Tim Cook has said rising memory costs mean handset prices will be going up too.

That comes via an interview with Apple's CEO in the Wall Street Journal, in which Cook said memory chip prices had become "unsustainable."

RAM prices have skyrocketed because of a drive for data centre demand sparked by AI, with prices of some models more than tripling by the end of last year. IDC said at the time that it was "the end of an era of cheap, abundant memory and storage". The data centre demand has led manufacturers to shift to equipment to meet those needs, away from components for consumer-grade hardware, further exacerbating the problem, while constricted supply has meant prices of remaining parts have continued to rise – not helped by a rise in scalper bots targeting what remains.

Amid these challenges, it's perhaps no surprise hardware prices are set to rise – and no surprise that Cook is concerned, as Gartner said earlier this year that PC shipments were at their lowest level in a decade.

"Unfortunately, price increases are unavoidable," Cook said, according to the WSJ. "We're doing our best ⁠to mitigate the huge increases that are being passed to us, and we've been trying ​to shield our customers from the increases, but the situation has become unsustainable."

How much will prices rise?

Cook didn't get into details, however, so it's unclear when those price rises will kick in, or which products they will impact.

But he did pin the blame on AI, saying the drive for higher-spec memory was impacting consumers. "There's less supply at a time when consumers want devices and the memory guys are passing along huge price increases," he said. "We definitely need memory pricing and supply to return to reasonable levels for ​consumer products. That's the ​bottom line."

Chiew Le Xuan, an analyst at Omdia, told the BBC that the average price of smartphones will increase by 20% this year, predicting that iPhones will cost $150 more this year – though that will also come with upgrade specifications. He added: "This is the new pricing reality, not a temporary spike."

What can be done?

Asked if the US should ease controls around sourcing such parts from China, Cook said: "Everything needs to be on the table."

Cook also suggested Apple was willing to help fund solutions from its own massive cash pile – which currently sits above $68bn, with total liquid reserves above $162bn – though it has no plans to build its own memory manufacturing. "We're willing to use our balance sheet ⁠to ​help be a part of the solution," he said. "Obviously, ​more capacity is needed. "

The BBC reported that Trump claimed on his social media site that Apple would be working with Intel to produce chips in the US, but any such deal has yet to be confirmed by either company. Last year, the US government bought 10% of struggling Intel as part of a bid to bring chip manufacturing to the US.

Cook is set to step down as CEO in September, handing the reins to John Ternus.

Apple has announced a long-awaited overhaul of its Siri personal assistant in a move analysts have described as a clear signal of its future AI approach.

As part of the upgrade, Siri will be integrated with Apple Intelligence, the tech giant’s in-house AI platform, and will be renamed ‘Siri AI’.

Expected to launch later this year, the upgrade will reframe Siri as more of a chatbot akin to the likes of ChatGPT, Claude, or Google Gemini – the latter of which underpins the AI assistant following a billion dollar deal.

This means Siri will provide users with answers to queries sourced from information on the web in addition to traditional personal context-focused queries.

Apple described the revamped Siri as “profoundly more intelligent, knowledgeable, and capable” compared to its predecessor.

Siri revamp highlights Apple’s AI focus

The move by Apple follows a lengthy process to deliver functional AI capabilities for users, an aspect of the company’s approach that has drawn criticism from customers and raised questions over its direction with the technology.

While Apple touted AI-powered upgrades to the assistant back in 2024, the company has repeatedly postponed any concrete announcements on this front. Indeed, the company appeared to acknowledge its lack of progress on this front at its Worldwide Developers Conference yesterday.

Craig Federighi, senior vice president for software engineering at Apple, said the company has taken a measured approach to integration of AI across its product ranges.

“Some appear to be racing forward, seemingly pursuing AI for the sake of AI, without clear regard for the people, all of us, that it’s ultimately meant to serve,” he said.

Ben Wood, Chief Analyst at CCS Insight, said Apple provided users with “some answers” to long-running questions about its shortcomings in AI.

Crucially, while new features outlined by the company aren’t exactly groundbreaking, the company’s track record in streamlining integration could make a big difference.

“Having acknowledged its shortcomings at last year’s event, there was no margin for error,” he said. “Many of the new AI-powered capabilities are similar to those offered by rivals, but Apple has a strong history of integrating such features in ways that make them more accessible to users across its deeply integrated hardware, software, and services.”

“It is also clearly seeking to differentiate through its privacy promises,” Wood added. “This looks like a step in the right direction, but there is no room for complacency, and Apple still has a long AI journey ahead.”

On Gemini’s involvement in the Siri upgrade, Wood said that speculation had been mounting ahead of the event over whether any upgrade would essentially be “Apple Intelligence powered by Google Gemini”.

Notably, however, the tech giant appears to have used the model to “accelerate and improve” the development of its own foundation model – a crucial factor for the company when it comes to privacy-related considerations.

“Apple remains in charge of pre-training, post-training, and reinforcement learning and is putting a strict privacy layer around the entire experience,” he said.

Forrester VP principal analyst Dipanjan Chatterjee echoed Wood’s comments both on privacy and potential integration challenges. Apple’s unique selling point lies in simplicity, he noted.

“In classic Apple fashion, the focus shifts from underlying technology to outcomes: usefulness, simplicity, and trust. The lesson for brands is clear: market the value, not the ingredients,” he said.

Winning over skeptics

While the overhaul marks a “watershed moment” for Apple, Forrester VP principal analyst Dipanjan Chatterjee said the company’s stop-start approach to AI over the last three years means it still has many skeptics to win over.

“After stumbling with the Apple Intelligence roll-out, Apple’s success will hinge on delivering the new Siri experience quickly – and ensuring it works as promised for iPhone users at scale,” he commented.

“I think Apple will convince the only skeptics that really matter – it's customers and prospects who may be getting their AI fix elsewhere – so long as they deliver what they have promised, and in short order.”

Chatterjee noted that Apple customers have “cut them plenty of slack” so far despite obvious shortcomings. Indeed, the company hasn’t witnessed a negative effect on revenue growth or its user base, but that could change if the rollout once again stalls.

“That grace period has an expiry date. A second miss, after Apple Intelligence, with Siri AI could hasten that date significantly.”

FOLLOW US ON SOCIAL MEDIA

Dell has announced the launch of the new XPS 13 laptop range as the company targets “accessible” pricing options for customers – and sets the stage for a showdown with the MacBook Neo.

The move builds on a commitment by the tech giant at CES 2026, namely competing “at every price point” in the hardware market.

Indeed, pricing for the XPS 13 comes in at $699, although a student-focused promotional campaign will bring this down to $599 in July - that puts the device on-par with the Neo.

According to Dell, the XPS 13 is the “lightest and most accessible” XPS to date, and it’s also the lightest and thinnest. The laptop measures 12.7mm and weighs only 1kg (2.2lbs).

The company specifically highlighted the fact the device is smaller and half a pound lighter than the MacBook Neo, and also boasts a larger display.

“The lightweight and durable chassis is built to the same standard as every XPS, because a lower price shouldn’t mean a lesser build,” the company said in an announcement.

So what can customers expect with the XPS 13?

Under the hood of the Dell XPS 13

The base version of Dell’s new budget laptop will come equipped with a six-core Intel Core 5 320 “Wildcat Lake” chip.

Dell noted that future versions will also be available with Intel Ultra Core “Panther Lake” chips, although information on the timeline here is yet to be revealed.

The XPS 13 also boasts 512GB of storage, and just 8GB of RAM for the base version, with the option of scaling up to 32GB.

That might seem low, but the MacBook Neo also comes with what some would describe as paltry RAM capabilities.

Elsewhere, the XPS 13 comes equipped with two USB-C ports, but no 3.5mm audio jack.

XPS 13: Design and features

The base version – and future versions – of the XPS 13 will come with a 13.4 inch touchscreen as standard, boasting 2560 x 1600 resolution and 30Hz to 120Hz variable refresh rate for “smooth visuals”.

Dell noted that the device also has a streaming battery life of up to 17 hours, largely due to the fact it automatically adjusts refresh rates based on what’s on the screen.

Other key features include:

• A 2.5K touch display (covering 100% of the DCI-P3 color gamut)
• Backlit keyboard
• Faster USB-C (3.2 Gen 2)
• Intel Wi-Fi 7
• Windows Hello
• Quad speakers

A showdown with Apple

Dell is keen to emphasize the XPS 13’s comparison with the MacBook Neo – and distinct advantages. Indeed, the company’s blog post referenced the device four times.

“Apple’s MacBook Neo is a capable machine, and its arrival confirms that there’s real appetite for premium quality at accessible prices,” the company said. “Where Dell differs is what we think premium means at this price point and what we were willing to build to deliver it.”

The MacBook Neo took the industry by storm earlier this year, generating huge interest and marking the company’s first foray into the budget device market. As ITPro noted at the time, the MacBook Neo launch set the stage for the rest of the industry to react on pricing.

Pound for pound, the two shape up fairly similar in terms of storage and RAM options. The XPS 13 entry-level device offers 512GB storage and 8GB of RAM, while the Neo comes equipped with 256GB storage and 8GB of RAM.

It’s worth noting, however, that due to Apple using in-house developed parts - chips, operating systems, and hard drive - the Neo could eke out more performance from just 8GB compared to a comparable Windows device.

From a visual perspective, the situation is much the same. The MacBook Neo is Apple’s smallest laptop to date, boasting a 13in display and coming in at just 12.7mm thick. The XPS 13, meanwhile, is 12.7mm thick with a slightly larger screen.

Who’s this for?

Dell specifically highlighted students as a key demographic for the XPS 13. However, much like the MacBook Neo, the pricing could be tantalizing for smaller businesses aiming to get more bang for their buck.

The launch of the Neo came at the height of the RAM crisis with prices spiralling globally, and lingering concerns about hardware costs mean Dell has an opportunity to capitalize in the budget market.

Essentially, small businesses now have another viable option alongside the Neo – and that's massive news given low budget devices were expected to be among those worst hit by the memory crisis.

FOLLOW US ON SOCIAL MEDIA

April has come to an end and what a busy month it’s been, with major announcements, updates, and cancellations across all corners of the tech sector.

Earlier this month, OpenAI made headlines throughout the UK with the news that it was pausing its landmark Stargate UK project citing region-specific difficulties. But how do these claims stack up, given signs OpenAI is pulling back from other key compute projects?

Also this month, John Ternus was announced as the incoming CEO at Apple. What does the Apple’s hardware king mean for the tech giant going forward? And DeepSeek released its much-awaited v4 Pro frontier model – with a muted response.

In this episode, Rory is once again joined by Ross Kelly, ITPro’s news and analysis editor, to discuss some of April's biggest news.

Highlights

"Stargate UK was this flagship expansion to compute capacity across the UK. Big money involved, big names involved. But now it's hit the rocks, largely due to a combination of sort of factors that, I suppose, are outside of OpenAI's control."

"So John Ternus is an apple veteran. He was the SVP for hardware engineering at Apple, he's been there since 2001, he's part of the bricks and mortar of Apple. It makes perfect sense that you would have Ternus move into this role. "

"We are edging closer to companies like deep seek being able to completely rely on Chinese energy production, Chinese chip production and Chinese AI training, essentially making them completely resilient to any shifts, bans, tariffs, embargoes that the US may experience or want to impose."

Footnotes

• OpenAI hits the brakes on Stargate UK infrastructure project, citing energy costs and regulatory concerns
• Has another OpenAI Stargate project hit the rocks?
• OpenAI announces first permanent London office in UK expansion push
• MacBook Neo: everything you need to know about Apple's £599 laptop
• M5 MacBook Pro is a minor spec bump with increased GPU performance – but once again, the immense battery life is the standout feature
• The Apple Mac Mini M4 is an affordable powerhouse that's perfect for any office desk – and it's also utterly adorable
• Who is John Ternus?
• ‘Steve Jobs was never going to be an easy act to follow’: How Tim Cook transformed Apple into a ‘durable’ tech industry powerhouse
• DeepSeek V4 Preview Release
• DeepSeek rocked Silicon Valley in January 2025 – one year on it looks set to shake things up again with a powerful new model release
• Using DeepSeek at work is like ‘printing out and handing over your confidential information’

Subscribe 

• Subscribe to The IT Pro Podcast on Apple Podcasts
• Subscribe to The IT Pro Podcast on Spotify
• Subscribe to the IT Pro newsletter
• Join us on LinkedIn

Apple CEO Tim Cook has announced plans to step down after more than 15 years leading the company.

John Ternus, head of hardware engineering at the tech giant, will assume the role of chief executive on 1 September, with Cook stepping into the role of executive chairman.

Cook will work closely with Ternus during the transition period, and will continue to “assist with certain aspects of the company” in his new role as executive chairman.

In a statement, Cook said serving as CEO has been the “greatest privilege of my life”.

"I love Apple with all of my being, and I am so grateful to have had the opportunity to work with a team of such ingenious, innovative, creative, and deeply caring people who have been unwavering in their dedication to enriching the lives of our customers and creating the best products and services in the world,” he said.

Cook’s replacement is an Apple veteran, having spent more than 25 years at the company. Ternus currently leads the hardware engineering division and has long been touted as a leading figure to replace Cook.

Cook added that Ternus “has the mind of an engineer” and is well-placed to take the reins at the company.

“He is a visionary whose contributions to Apple over 25 years are already too numerous to count, and he is without question the right person to lead Apple into the future,” Cook commented.

“I could not be more confident in his abilities and his character, and I look forward to working closely with him on this transition and in my new role as executive chairman."

Tim Cook had a hard act to follow

Cook first joined Apple in 1998 and has served as CEO since 2011, having stepped in to replace co-founder Steve Jobs just months before his death.

Under his leadership, the company has expanded significantly, launching a range of new products including the Apple Watch, Apple Vision Pro, and AirPods series - with Ternus closely involved in the development of these ranges.

Services including iCloud, Apple TV, Apple Pay, and Apple Music were also launched under Cook’s leadership. The iPhone, a staple of the company’s offerings, has retained continued popularity across his tenure.

In 2018, the firm became the first public to be valued at $1 trillion. Since taking over, the company’s market capitalization surged from roughly $350 billion to around $4 trillion.

Cook hasn’t quite achieved the same status as a tech visionary like Jobs, however.

Dipanjan Chatterjee, VP principal analyst at Forrester, said Jobs “was never going to be an easy act to follow”, yet despite this the Apple chief transformed the company into a “durable, resilient financial powerhouse”.

“Cook’s legacy will be defined by steady, disciplined operational stewardship - proof that a company can be more than just exciting and visionary; it can also be immensely valuable to all its stakeholders,” he commented.

Chatterjee noted that while Apple’s growth trajectory has moved at a steady pace over the last 15 years, Cook hasn’t exactly overseen a step-change in innovation that would “reset Apple’s competitive position for the next two decades” – much like Jobs did with the iPhone.

This means the tech giant is still largely “structurally dependent” on the product range and requires a new growth engine.

Since the advent of generative AI in late 2022, questions have been raised about Apple’s approach to the technology, with some industry stakeholders highlighting its tentative steps in this domain.

“That’s where John Ternus comes in,” Chatterjee said. “Ternus is a hardware engineer, which signals that Apple will seek differentiation in its physical products even as it looks to reframe the device as a substrate for intelligent experiences.”

Chatterjee added that Ternus’ appointment could mark a “quiet pivot back toward product intimacy” as well as a tighter coupling between hardware, software, and future AI capabilities.

FOLLOW US ON SOCIAL MEDIA

As legend has it, to finance the development of the Apple I personal computer kit Steve Jobs sold a Volkswagen bus and Steve Wozniak an HP-65 calculator. This is how Apple was born, in the garage of a Californian house on April Fool’s Day 1976. Fast forward to today, and the company is celebrating its 50th anniversary.

Reportedly, Jobs eventually replaced the bus with supercars as Apple rapidly grew into a consumer tech giant, making him a millionaire in his early 20s. That fortune was made through the consistent development of groundbreaking products – a feat Apple has continued long after his passing in 2011.

And so, today, we can look back through five decades of innovation – across computing hardware, software, mobile devices, and MP3 players – and see a company that has made creating culturally significant products its business. Here, the ITPro team talks about their own favourite Apple products.

This is a really hard choice for me, as I’ve had so many Apple products in my life – many or even most of which I have enjoyed. I’m going to go back to where it all began, though, and pick the Apple Macintosh Performa 630. This desktop computer arrived in my childhood home in about 1994, taking up residence between the dining table and the window to the back garden. Until that point, my main interaction with computers had been at school using BBC Micros and perhaps the odd Acorn Archimedes, so the Performa 630 was quite the revelation.

The Macintosh System Software (as the Macintosh OS was called at the time) GUI was a big change compared to even the Windows PCs I’d come across, let alone the school computers. It proved an entry point for word processing and databases, courtesy of Claris Works, as well as desktop publishing via QuarkXpress. Sitting at my iMac 32 years later, I can’t help but think this may have helped steer my adult life in a certain direction…

I’m in a bit of a different position than my colleagues, having come very late to the Apple ecosystem. Apart from the iPod Touch I was gifted in my early high school years – which was equal parts Temple Run and Linkin Park machine – I was a 100% Windows and Android user until coming to work at ITPro.

Much has changed in the intervening years. My reaction to being given a MacBook in my first corporate job was one of surprise and worry, which quickly gave way to joy as I realized the simple brilliance of macOS. Indeed, I increasingly found myself frustrated with the relative clunkiness and DIY reality of using Windows, and my opinion of it further soured with the switch to Windows 11.

All of this led me to purchase an M4 Mac Mini at the end of 2025. I’d previously had the opportunity to review the M2 Mac Mini for ITPro and had been thoroughly charmed by the small but mighty output of this pint-sized device. My partner can attest to the numerous times that, wandering around an Apple store while out in Central London, I would linger at the Mac Mini section and comment on just how little seems to faze it.

Having now adopted the M4 variant, which is even smaller than its predecessors, there’s no looking back for me. In my personal time, I love photography and video editing, and so far I’ve found the Mac Mini can handle software like DXO PhotoLab and DaVinci Resolve with no grumbles (literally – it’s eerily silent).

Recent crazes like OpenClaw have run down stocks of the Mac Mini, as AI enthusiasts look to put the affordable device to work automating away their daily tasks. But if you can get your hands on one and use it manually, you won’t look back either.

My family home was most certainly an Apple shop as a teenager, and still is as I approach my mid-30s. My mother couldn’t be separated from her iconic U2-branded iPod, and my brother and I were seldom found without our iPod Nanos - in different colors to avoid mix-ups, of course.

With the iPhone, the situation is no different. My first proper smartphone was the iPhone 3GS, and it felt like a taste of the future. No more clunky buttons, just a sleek touchscreen device that had everything you needed: calls, texts, music, it was all wrapped up in one.

During the launch ceremony for the device in 2009, Phil Schiller, Apple’s SVP of worldwide product marketing, noted the ‘S’ stood for ‘Speed’ - and he wasn’t wrong.

The device was markedly faster than its predecessors, up to twice the speed, and boasted a longer battery life. Storage options (16GB and 32GB) were also fantastic for the time.

Looking back, the iPhone 3GS set a high bar for smartphone performance, and while it pales in comparison to what the company offers now it certainly was a pioneering device.

Performance was only one part of the equation though. Ease of use, the built-in camera, and the convenience of downloading whatever app you wanted through the App Store made it the perfect digital companion.

A brief period with a BlackBerry was enough to convince me that iPhones were the only viable option when it came to a mobile device. From the 3GS to the brick-like 4G, right through to my current iPhone, the 15 Pro, I’ve been all-in on Apple, and it won’t change any time soon.

As the ITPro reviews editor, I’ve had the privilege to review many of its laptops and mobile devices. But for me, it is the consumer products that leave a bigger mark on the world.

Steve Jobs didn’t promise to put “1,000 songs in your pocket” to his daughter; that’s just sentimental fiction from the 2015 Danny Boyle film about Jobs. He did, however, say the line on stage during the announcement for the first iPod in 2001. The slogan was powerful as a great marketing gimmick and a somewhat skewed truth in the legacy of Jobs. I quickly burned all my favourite songs into the original iPod. I’ve gone on to use the Shuffle, the Nano, and, of course, many iPhones in my time. Often taking for granted the sheer innovation behind having my whole music library with me, wherever I go.

The US comedian Bill Burr has a funny line on Jobs and the iPod. It plays on the fact that Jobs famously wasn’t a coder or a technician. He merely delegated ideas to people. So at some point during the new millennium, Jobs turned to Apple engineers, led by Jon Rubinstein, and told them to build a device that could fit in your pocket but hold all your songs. Reportedly, size was the key. There’s a story about Jobs plunging a prototype into a fish tank to show that bubbles emerging from the device highlighted space that could be removed.

Nothing, for me, quite highlights the lore of Jobs at Apple like the iPod. But also, given that this feat only took a year and comprehensively changed consumption of music, film, TV, and culture, it’s an example of Apple at its very best.

The Apple iPhone 17e is, in my humble opinion, an example of what the company does best. It takes its premium wears and strips them back into an affordable, yet innovative offering – see the MacBook Neo. This is a tightrope walk of leaving you with just enough of what you need and dialing back, where possible, on flash. For phones, that means fewer camera lenses and perhaps a duller screen.

However, making room for the A19 chip has brought a world of innovative features to a modest budget. Add some clever buttons, MagSafe connection, and an almost nostalgic design, and you have an alarmingly attractive mid-range smartphone.

At $599 (£599), the iPhone 17e is considerably cheaper than the lowest Galaxy S26 device. And, going by the price of the S25 FE, it might endup being cheaper than the S26 FE (should there be one released later in the year).

Apple iPhone 17e: Design

The easiest way to spot that this handset is the 'e' model of iPhone 17 is the single camera lens. I'm now at a point in life where that is an odd sight on a phone, as it's normal for them to have two and three (or more). So the rear of the phone looks a little old-fashioned. The rest of the design is more akin to the iPhone 14; really compact, with flat aluminum edges. It fits nicely in the hand, in the same way that the iPhone 12 Mini did five generations ago.

Another noticeable difference is the notch on the display, which is considerably longer than on the standard 17. However, said display is tougher with the new ceramic compound and the latest Gorilla Glass shielding.

Other physical additions include the Action Button (above the volume ones), which I instantly loved. I like the name and its function(s). You can choose from a collection of different shortcuts, like the camera or the focus app – only one action can be used, however. You can use files and have the action button bring up a maximum of four apps/tools, but that makes it more like the drop-down software menu. Admittedly, the button could be even better with a double or triple-click function for extras. I used it for the camera initially, but I have also switched it to the torch. Why can't I have both?

Apple iPhone 17e: Display

You get a 6.1in Super Retina XDR OLED display on the 17e, with a 2,532 x 1170 resolution. This is both far bigger than the 4.7in iPhone SE, and has a greater screen-to-body ratio than the iPhone 11.

You don't get a 120Hz refresh rate or an always-on display, but that's the only real omission here. You still get True Tone, Wider Color support, and a very bright panel. I recorded peak brightness at 535 cd/m2 with our colormitere.

For almost every use (streaming, photography, web browsing), you get a great show of color. It's punchy when it needs to be, the blacks are deep and inky, and it has strong contrast. In truth, I didn't really notice the low refresh rate as it was still very smooth when scrolling the socials.

We also recorded 97% for sRGB gamut color coverage, which backs our experience of using the 17e. It isn't a groundbreaking device, but for a budget device, the display is great value for money. The larger-than-normal notch was also not that big an issue, despite cutting into the screen space.

Apple iPhone 17e: Specs and Performance

Having the A19 chip in the iPhone 17e is a bit of a cheat code. Although it's a slightly different version from what is found in the iPhone 17, it's still a 3nm chip with a 6-core CPU, 4-core GPU, and a 16-core neural engine. And that gives this budget phone a lot of premium features and operations. You also get double the storage compared to the 16e (128GB to 256GB).

In Geekbench 6, the 17e scored 3,648 for single core and 9,215 for multicore. Both benchmarks are a significant improvement over the Apple iPhone 16e, but also pretty close to the standard iPhone 17, which scored slightly above for multi-core.

The doubled storage also gives you plenty of room for apps, music, film downloads, and photos. Or even videos captured with the 48MP main camera – I was able to shoot and store enough for two videos.

In day-to-day use, I found the 17e to be lightning fast, responsive, and capable of getting me through my work day and beyond. It did seem to get quite warm on the back when downloading lots of apps in a row, during the setup, but that was the only time that issue came up.

In our looped video test, the 17e lasted 19hrs and 13mins, which is exceptional for a phone of this price. Again, having the A19 chip and its AI capabilities helps with optimization. In real terms, the phone lasted through work days, though it did noticeably drain quicker on days I shot video and transferred a high volume of files.

Apple iPhone 17e: Cameras

You get one camera lens on the rear of the iPhone 17e, but with two settings; it's a 48MP 'Fusion' camera, as Apple calls it. The additional capability is a 2x telephoto zoom. It's also capable of 4K video (fps?)

There are a couple of annoyances here; it's a 4x3 aspect ratio by default, but can be switched to 16x9 or 1x1. Images don't quite capture what you see before you hit the shutter button; they come out a little overexposed. And then there is a delayed enhancement that's particularly noticeable when you take selfies, which is a bit hit and miss. You also don't get a 'Pro' mode to tweak shutter speed or ISO. There's just a toggle for 'Exposure' that didn't seem to make much difference to the final shot.

Oddly, the video capabilities are much better. You can record in 4K at 60 fps, and it has a range of automatic features that help you, such as the wind noise reduction feature. You can also film multiple scenes in one, using the pause and record feature. We shot our most recent YouTube shorts on the iPhone 17e, and you can see the quality for yourself.

Apple iPhone 17e: Is it worth it?

The iPhone 17e seems like an internal challenge to Apple; how many of the most converted features can you fit into a budget handset? The answer this year is a lot. So you get an A19 chip and all its computational benefits. MagSafe. A pretty innovative camera system, with great video capabilities, and a rather long-lasting battery.

One could argue that the camera isn't high-quality enough – Honor and Samsung are offering a lot more (lenses and software) for roughly the same price. But the overall experience of the iPhone 17e is superior, in our opinion. It comes down to what you want from your budget phone. If it's a top-of-the-range camera with video capabilities, you would surely look at the premium models anyway.

For those who want a cheaper way into the Apple ecosystem, or businesses that want to find quality handsets that their employees will actually like, the iPhone 17e is undeniably a solid choice.

Apple iPhone 17e specifications

CPU A19 6-core CPU, 4-core GPU, 16-core neural engine

Display 6.1in Super Retina XDR OLED display

RAM 8GB

Storage 256GB or 512GB

Connectivity Wi-Fi 6, Bluetooth 5.3

Cameras (Rear) 48MP Fusion Camera System (Front) 12MP

Dimensions (DHW) 7.8 x 146.7 x 71.5mm

Weight 170g

Operating system iOS 26

Apple has announced a new all-in-one platform for businesses that allows them to manage apps and devices, get expert support, and grow their business.

Apple Business, which launches in April, replaces the company’s current business portfolio (Apple Business Essentials, Apple Business Manager, and Apple Business Connect) and rolls them into one simplified platform.

Customers who join the plan will get access to features such as ‘Zero Touch’ deployment, where devices can be sent straight to the user with next to no input from an administrator.

Auto enrolments help to onboard new employees and keep existing staff working when a new machine is needed. Businesses can also customize their Apple email, calendar, and directory with their own branding.

“Apple Business is a significant leap forward in our decades-long commitment to helping companies of all sizes leverage the power of Apple products and services to run and grow,” said Susan Prescott, Apple’s vice president of Enterprise and Education Marketing.

“We’ve unified Apple’s strongest business offerings into one simple, secure platform, delivering key features for organizations in every stage and sector, including built-in device management, collaboration tools, and additional ways to reach new customers.”

Apple Business: Run and Grow explained

The new platform is split into two distinct sections: Run and Grow. The former is largely concerned with the management and maintenance of devices and software. Grow, meanwhile, offers some unique capabilities to aid businesses with marketing and visibility.

A key element of this is Apple Maps, which will soon allow businesses to create their own ads inside the application.

Businesses will be able to use automated tools to create their own ads and also improve how they appear in Maps with a set of brand management tools.

Many of these will be tools previously found in Business Connect, though there will be analytical functions to see how businesses are discovered through Maps.

Apple Business will be available from 14 April as a free service in the US and largely free in over 200 countries around the world.

For US customers, there is a fee for additional storage (up to 2TB), which starts at $0.99 per user per month. However, there are no additional storage options for customers outside of the US.

The launch of the new service will mean that Apple Business Essentials, Apple Business Manager, and Apple Business Connect will all be shut down, with existing Business Connect customers automatically migrated to Apple Business by the launch date.

FOLLOW US ON SOCIAL MEDIA

The secret to great product launches appears to be the same as comedy (Timing). And so the Apple MacBook Neo, the $600 (£600) notebook that was instantly ubiquitous across social media, comes to us at the end of the first quarter of the year. A year likely to be dominated by a looming shortage of RAM and a potential hike in the cost of all hardware.

It's got just 8GB of RAM, uses Apple's excess stock of A18 mobile chips, and comes in a range of appealing colors. This is all to the delight of podcasters and influencers who talk about this type of stuff. Is it underpowered? Is it overpowered? Is this going to fundamentally change our expectations of entry-level laptops? We're going to find out with this review...

Apple MacBook Neo: Design

The MacBook Neo is unmistakably a MacBook; the lid, the logo, the bezels, the keyboard, it's all there. But there are a number of cutbacks, both inside and outside, that make it so cheap (for Apple). There's no backlighting for the keyboard, and you only get two USB-C ports and a headphone jack (all down the left-hand side). Both ports take charge, but only one hooks up to an external monitor.

This is the smallest MacBook available with just a 13in display. It's only 12.7mm thick and 297.5mm wide. Noticeably less than both the M5 models of Air and Pro. The keyboard is still full-sized, with largish keys, though the trackpad is a little on the small side.

A lot is being made of the colors, and there is a huge slice of nostalgia about them. You get four options here: Citrus, Blush, Indigo, and Silver. These are all nice, though some are more professional than others (we have the Silver). Apple devices that go on to be iconic are often bright colors; think iBooks or the orange iPhone 17 Pro.

Apple MacBook Neo: Display and keyboard

For a $600 laptop, the Neo has a surprisingly bright panel. It's a 13in IPS display, rather than OLED, but I recorded its peak brightness at 523cd/m2. Which is more than enough to be readable in most conditions, and I can't say I ever struggled to see the screen or needed to adjust its levels.

Naturally, the lack of an OLED display will lead some to suggest the Neo's display is not up to scratch. But in our tests, the IPS display came back with a strong show of color reproduction. It isn't so great with HDR videos, but its general quality is high enough for most users. I found it perfectly capable for editing images, streaming, and working late at night. In short, this is one of the best budget displays around.

The first thing to say about the keyboard is that it's worth paying extra to have the Touch ID. I know that's $100 (£100) extra, but it's very frustrating to have to type out a password every time you open it or the screen shuts off. And the lack of backlighting might be a problem for first-time laptop users at night, like young students. The rest of us type without looking, so it shouldn't be an issue.

Beyond that, the keyboard is just as good as you get on a Pro or Air; nice punchy keys with plenty of travel. The clack sound it makes is soothing. And you get a full-sized function row.

The trackpad is also just as good, or nearly the same as you get on the premium models, though it is noticeably smaller. Multi-touch gestures make the whole operation so simple – four finger swipes to see your desktop – both a great intro to MacBook but also the same experience users will know and love.

Apple MacBook Neo: Specs and performance

The MacBook Neo is a great example of mobile chips in laptops; it features the A18 Pro processor, which was used in the iPhone 16 Pro. This is a chip purposely built for on-device AI workloads, and with it, a range of nifty optimizations across the device. Having a laptop with all in-house developed parts (chip, hard drive, operating system) allows Apple to get the most out of it. So the 8GB of RAM isn't that big an issue, like it might be on a similarly priced Windows machine.

In Geekbench 6, the Neo came back with 3,287 for single-core workloads – the A18 Pro is able to give one application all it needs (and requires), which makes for a smooth operation. For single-core scores, the Neo is actually much stronger than MacBooks with the M1 and M2 chips. The multi-core score is also good, but 7,804 is far below everything from M2 onwards.

On the question of whether the Neo is either over- or underpowered, in our experience, it's somewhere in the middle. Editing images in Photoshop, for example, was perfectly fine. And, to some degree, editing video was OK. There's buffering. You'll see that little rainbow wheel frequently, but briefly. Sometimes you'll drag a clip, and it will disappear. Once you release the click, it comes back, but it's a little frustrating. I still made a YouTube Short out of 4 minutes-worth of footage. The Neo can do these things, but full professional video editing should be left to the Pro.

In our looped video test, the Neo lasted 13 hours and 18 minutes, which is both good for a £600 laptop and also great for remote workers (or those who spend lots of time away from plug sockets). This little wonder can get you through a day of work – even during our heavy video editing.

Apple MacBook Neo: Is it worth it?

The most disruptive product Apple has released since the iPhone, but it might not be disruptive to Windows machines, as there's too much choice within that bracket. It might, however, make a sizable dent in the Chromebook market.

In the next couple of years, you may go to IT for a temporary machine, or an entry-level one, depending on your role, and you'll likely see that old cupboard of unwanted Chromebooks replaced by Neos.

Apple MacBook Neo specifications

Apple made waves at the start of March with the launch of the MacBook Neo, its first budget laptop.

As RAM prices skyrocket, businesses are under pressure to make device purchases that count and that will future-proof their workers for years to come. Both Apple and manufactures in the Windows ecosystem are vying for enterprise money in the budget and pro segments – even as it becomes more expensive than ever to spec devices to 16GB of memory and beyond.

Where does the latest Apple release fit into the wider business laptop ecosystem?

This week, Rory speaks to Bobby Hellard, ITPro’s reviews editor, to unpack all things MacBook Neo, Windows laptops, and the future of business hardware.

Highlights

"It's just a clever kind of construct. It's got a mobile chip, very low RAM, the display isn't as the best, and it's got some things missing from it that you would find on the MacBook Pro or the MacBook Air. So it's for people that want a MacBook but don't want to have the best of the best, don't want to spend so much money."

"So for single tests and single apps, it's brilliant, and that's part of the reason why they've got that A18 chip in there. It's an AI ready chip, and it can optimize the laptop better than it would with an M1 chip. So you get enough power and specific power to various apps so they work perfectly, no more, no less.

"Well, it felt like a return to form for MWC this year, because there were a lot of handsets, a lot of showcases. There were some foldable phones, folding screen laptops as well – Huawei was one of those it has, which we've reviewed, the MateBook Fold, which is an all-display laptop that folds. That was on display, it's very beautiful. Lenovo had some modular PCs, so they've got a new modular concept, which is an AI PC that you can completely tear down and pull apart, which has loads of use cases, particularly for fixing and sustainability."

Footnotes

• MacBook Neo: everything you need to know about Apple's £599 laptop
• The MacBook Neo is another Apple master stroke, and it couldn’t have come at a better time
• Low-budget devices are the biggest casualty of the RAM crisis
• Give businesses more practical AI services and some return on investment before you go selling 6G
• Lenovo wows at MWC with concepts for a modular ThinkBook and desktop AI-devices
• AMD thinks ‘AI on the PC has crossed an important line’ – the rise of ‘Agent Computers’ will be the next big breakthrough
• Are AI PCs becoming the norm?
• Nobody asked for AI PCs, but they’ve been thrust upon us – at least Dell realizes it’s probably made a mistake

Subscribe 

• Subscribe to The IT Pro Podcast on Apple Podcasts
• Subscribe to The IT Pro Podcast on Spotify
• Subscribe to the IT Pro newsletter
• Join us on LinkedIn

Apple’s MacBook Neo announcement sent shockwaves across the hardware industry last week, and the arrival of a budget device from the tech giant couldn’t have come at a better time.

The launch of a $599 (£599 in the UK) laptop is a bold move by Apple, enabling it to tap into demographics previously excluded by the high prices associated with its flagship models.

The 14in MacBook Pro M5 comes in at $1,699 (£1,699 in the UK), while the MacBook Air M5 will set you back $1,099 (£1,099 in the UK). That’s a significant price drop.

At the very least, the MacBook Neo will force the rest of the industry to react. This isn’t so much a race to the bottom on cost, but a sharpening of the focus on giving consumers better bang for their buck.

Some, it seems, are already at panic stations over the announcement. Speaking during a recent earnings call, Asus’ chief financial officer Nick Wu said competitors will have to react swiftly or be left in the dust.

“Given Apple’s historically very premium pricing, launching such an affordable product is certainly a shock to the entire market,” he said, per reports from Seeking Alpha.

The fact this has come as such a shock to the market speaks volumes about its recent trajectory. The incessant focus on AI-powered devices, which come with a premium price tag, is an issue I’ve raised in previous columns for ITPro.

We’ve seen recurring Windows-related gripes, with users pushing back against efforts to transform the operating system into an “agentic OS”. This sentiment stacks against Apple’s competitors – even if the tech giant is itself embedding generative AI throughout its products via Apple Intelligence.

A viable option for small businesses

While Apple is primarily targeting the education sector and students with the MacBook Neo, there is certainly room here for small businesses to capitalize. IT leaders already sweating over costly device refresh programs and price hikes thanks to the RAM crisis will undoubtedly be considering the Neo.

The device will certainly offer better bang for buck. An array of comparison tests between the Neo and Windows-based alternatives already show there are glaring pros and cons.

Testing by ITPro found the Neo performs well in single-core performance, scoring a 3,287 in GeekBench 6. This puts it ahead of devices such as the Lenovo IdeaPad Slim 3x and the Dell 14 Plus – two devices ITPro has had hands-on experience with.

It’s once you get into multi-core tasks that issues arise, particularly when running multiple programs or specialist tasks such as video rendering. Tom’s Guide noted the Neo’s performance drops off here compared to its aforementioned competitors.

Apple has framed the Neo as a viable option as a work device. Promotional materials specifically touted its potential in enterprise domains.

“From building spreadsheets to managing payroll, MacBook Neo gets the job done,” the company claims. In many ways, the Neo does punch above its weight, but knowing the limitations and making them clear is an area Apple has been candid on in recent years - take its AI approach, for example.

The company has adopted a standoff-ish approach to AI over the last three years, hedging its bets on building a more reliable integrated AI assistant and making that strategy clear. It’s a tactic that’s drawn an equal mixture of praise and criticism from industry stakeholders, but one that’s ultimately paid dividends and underlines its long-running cautious approach.

Apple isn’t rash. It’s cold, calculated, and knows exactly when to strike.

The MacBook Neo is another Apple masterclass

Apple has a long-running reputation for shaking up markets; it did it with the iPod, the iPhone, the iPad, and the MacBook Air some 20 years ago.

This latest move may be the most consequential product launch in the company’s history – and given recent memory cost concerns, the impact cannot be understated.

Skyrocketing memory costs are severely affecting the budget device market, according to recent analysis from Omdia. These surging costs mean vendors are now focusing their attention on premium range devices to absorb the financial blow.

Omdia’s research specifically noted that PC prices below $500 will be hit the hardest, with global shipments plummeting by 28% across 2026. Simply put, you can kiss goodbye to your low-budget devices over the next year.

A device roughly within this price range, from a brand such as Apple, will be a tantalizing prospect for businesses and consumers alike.

MacBook Neo performance may be an issue

Performance may be a sticking point for some. But the benchmark situation is complicated and doesn’t paint a full picture.

On the surface, the MacBook Neo doesn’t quite shape up in terms of performance compared to its flagship products, or other Windows devices out there. The device will boast just 8GB of memory, with two distinct ranges offering 256GB and 512GB of storage.

Powering the Neo is the A18 Pro chip, an ARM-based system currently used in the iPhone 16 Pro and iPhone Pro Max.

Apple noted in its announcement that this is 50% faster at web browsing, while boasting three-times the performance handling workloads compared to PCs with Intel Core Ultra 5 chips.

An area in which the Neo has a distinct advantage, however, is the ecosystem it inhabits. Apple designs the interlinking parts under the hood, from the CPU and hard drive to the operating system.

Apple RAM also uses a unified memory architecture, creating a shared memory pool which allows the CPU and GPU to share bandwidth. What this means is that it better utilizes memory compared to a Windows device.

Add that to the broader Apple ecosystem and compatibility just makes sense given the demographic the company is targeting: students. Having your messages, photo albums, and various files accessible just adds to the ease of use factor.

Apple has unveiled the MacBook Neo, a laptop running on its iPhone chip that comes in hundreds of pounds cheaper than its siblings at $599 (and £599 in the UK) – perhaps opening up the laptop market, in particular in education.

The Apple MacBook Neo is powered by the A18 Pro, an ARM-based system on a chip that's currently seen in the iPhone 16 Pro and iPhone Pro Max. Apple claims that MacBook Neo's A18 Pro is 50% faster at web browsing and three times faster at on-device workloads than existing PCs with the latest Intel Core Ultra 5.

There's two storage configurations: the default 256GB, and a 512GB edition for an additional $100/£100. Both versions feature the same 8GB of memory – and there's no RAM upgrade possible. Apple's other MacBooks all start at 16GB.

Asymco analyst Horace Dediu said on social media that the Neo's arrival was a "watershed event".

"First Mac with a mobile processor and the end of the disruptive arc of mobile computing," Dediu wrote. "From Motorola to Intel to Apple silicon M, now personal computing is an accessory to mobile computing. A sharp punctuation point."

MacBook Neo price targets rivals

John Ternus, Apple's senior vice president of Hardware Engineering, said in a statement that the laptop had been "built from the ground up to be more affordable for even more people," adding that the MacBook Neo "delivers the magic of the Mac at a breakthrough price."

And it's the price that's raising eyebrows, and that's no surprise given Apple has traditionally positioned its MacBook lineup as a "premium computing product," noted IDC analyst Francisco Jeronimo.

"With such a product, Apple is aggressively aiming to grow market share while expanding the ecosystem," Jeronimo said. "Lowering the entry barrier to the Mac means Apple can bring more users into its services and device ecosystem, particularly students and first-time Mac buyers."

Indeed, the price point makes it a compelling option to draw into the Apple ecosystem any Microsoft users tired of Windows as well as a viable alternative to Google Chromebooks in education, helped by an additional $100/£100 discount for schools.

Ernest Wong, the head of research at Baskin Wealth, said on social media: "Every kid is going to want this to replace their Chromebook at school."

The MacBook Neo's price is a big step down compared to Apple's other laptops, though they are much more powerful in terms of performance. Alongside announcing the Neo, Apple also updated its MacBook Pro models. The 14in MacBook Pro with M5 now features 1TB of storage; prices start from $1,699 (£1,699 in the UK). The 14in MacBook Pro M5 Pro starts from $2,199 while the 14in MacBook Pro with M5 Max starts from $3,599.

A closer comparison is the MacBook Air, which now features the M5 in a 13in version with 16GB RAM starting from $1,099 (£1,099 in the UK) – but that's still a step up from the MacBook Neo.

MacBook Neo specs and availability

Beyond the aforementioned A18 chip, 256GB and 512GB storage, and 8GB of RAM, the MacBook Neo claims 16 hours of battery life, and weighs in at 1.2kg.

The MacBook Neo features a 13in Liquid Retina display, 1080p HD camera, and uses Apple's Magic Keyboard, with the multi-touch trackpad available in the 512GB edition that costs an extra £100. The MacBook Neo is available in four colours: blush, indigo, silver and "a fresh new" citrus; translated, those are pink, blue, silver and yellow.

Ternus said: "It features a durable aluminium design in four beautiful colours; a brilliant Liquid Retina display; Apple silicon-powered performance; all-day battery life; a high-quality camera, mics, and speakers; a Magic Keyboard and Multi-Touch trackpad; and the intuitive and powerful features of macOS."

IDC's Jeronimo said the timing is intriguing, given it comes at a time of memory and storage shortages causing price increases across the market – meaning an entry-level laptop featuring less RAM could be a smart move. "By using a chip that has only been used on the iPhone and the iPad, the real question is not whether Apple can sell a MacBook at this price (because it will be one of the most sold Macs ever if they can deliver), but how it balances cost, performance, and brand positioning while maintaining the premium experience that defines the Mac," he added.

MacBook Neo preorders are already open and will be available from next week. Apple also unveiled updates to its MacBook Air and MacBook Pro, as well as the iPad Air and introduced the iPhone i17e.

For some time, rumors have suggested that Apple is planning to release a touchscreen OLED MacBook. Reports from October stated the new device is due at the end of the year alongside other M6 devices.

If true, it’s likely Apple is already at a late stage of production on the devices. But there’s still time for the firm to kill the idea off, Google style.

Put simply, touchscreen laptops are a solution in search of a problem. For the most precise tasks, nothing beats a mouse or touchpad – and Apple already makes the best touchpads in the business. Touchscreens only make sense in a form factor where there’s no dedicated keyboard and mouse, which makes no sense for MacBooks.

It’s also hard to beat an iPad for sheer quality in the tablet space, though some of the best business tablets have Apple beat across specs and affordability.

Want a device for portable productivity? Buy a laptop. Want a larger touchscreen experience, with dedicated support for styluses and creative apps? Buy a tablet. The two should never meet.

More than anything, the inclusion of a touchscreen in a MacBook signals deep creative bankruptcy. It’s a true ‘we’ve run out of ideas’ move, which is weird considering it’s reportedly due to arrive alongside the long-requested OLED MacBook panel.

I can’t think of any functionality that it would offer, beyond niche use cases for the Apple pencil in creative apps – to which I would say, buy an iPad.

I greatly enjoy digital artwork in my free time, so part of me wants to love touchscreen laptops. Admittedly, I loved the Asus Zenbook 14 Flip OLED when I reviewed it in 2023 and praised the low latency of the display when paired with the official Asus Pen. But I know that over time, the downsides of a touchscreen display would have played on my mind.

Oh, the downsides. When you crack a standard laptop display, it can be fixed relatively easily and cheaply. Touchscreens are more expensive to replace, so I hope you have budget set aside if you’re equipping your entire workforce with touchscreen devices. When they do break, they can cause damage on their way out by registering random clicks all over the screen.

They’re also a total fingerprint magnet. If you thought your laptop screen was already a little grim – and be honest, it probably needs a wipe – I hope you enjoy the smears you’ll get from regular touchscreen usage.

Beyond it all, there’s no added value from having a touchscreen on your laptop. It’s added risk and cost for no real benefit, a misstep for any manufacturer but a downright sin if introduced to the MacBook lineup.

If Apple is out of ideas, it should see what can be done with e-ink. My inbox – and that of our reviews editor Bobby – is open.

Apple is turning to Google to help underpin its Siri service and “Apple Intelligence” capabilities, marking the latest sign the tech giant is gaining momentum in the AI race.

The duo have signed a multi-year partnership that will see Google support a major upgrade to Siri, which is expected to launch later this year.

Moving ahead, Apple’s foundation models will also be based on Google’s in-house Gemini AI model range, the company confirmed.

“After careful evaluation, Apple determined that Google's Al technology provides the most capable foundation for Apple Foundation Models and is excited about the innovative new experiences it will unlock for Apple users,” the duo said in a joint statement confirming the deal.

“Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple's industry-leading privacy standards.”

The move comes in the wake of several months of speculation over closer ties between the two tech giants. Google already pays Apple billions of dollars to remain the default search engine on iPhones, for example.

According to reports from Bloomberg in August last year, talks between the two over a potential deal for Gemini to power Siri were in the works. Separate reports from the publication noted the deal could see Apple pay Google upwards of £1 billion a year to use the model range.

The deal also represents somewhat of a snub for OpenAI. Apple previously integrated ChatGPT into macOS and forged close ties with the company on Apple Intelligence, for example.

Opting for a deal with Google suggests the tech giant views the Gemini range as a more reliable option than OpenAI's flagship GPT series.

Google is pulling ahead in the AI race

From Google’s perspective, the deal with Apple marks a huge seal of approval for its Gemini model range. Wedbush analyst Dan Ives said the partnership is a “major validation moment for Google” in a Monday research note, per reports from Fortune.

Alphabet, Google’s parent company, hit a major milestone in the wake of the deal, surpassing a $4 trillion valuation for the first time. The surge means Alphabet is just the fourth company to hit such a milestone alongside Nvidia, Microsoft, and Apple.

A key factor behind the breakneck growth at Google has been its growth strength in the generative AI race. The company has recorded consistent growth and progress on this front over the last three years following a rocky start.

When ChatGPT launched in November 2022, Google was caught off guard by the release, prompting a rapid shift in focus led by CEO Sundar Pichai.

As ITPro reported at the time, CEO Sundar Pichai issued a ‘code red’ alert which saw the company reallocate personnel and resources in a bid to keep pace with OpenAI and Microsoft.

Recent events suggest that Google has firmly caught up with key competitors in the AI space, particularly with regard to model capabilities.

In November, Google announced the launch of Gemini 3. The new model range outperformed OpenAI’s GPT-5 model across a range of key benchmarks. That launch precipitated a similar ‘code red’ scramble at OpenAI, with CEO Sam Altman pushing for the company to release GPT-5.2 in quick fashion.

A huge victory for Google

I wouldn’t be surprised if somewhere in the Googleplex, Alphabet’s sprawling Mountain View HQ, there isn’t a proud executive today hanging up a plaque that reads “Google's Al technology provides the most capable foundation for Apple Foundation Models”.

This is a major victory for Alphabet, particularly Google DeepMind and Google Cloud, which have spent the past year turning Google’s cloud empire into the foundations for AI dominance.

At Google Cloud Next 2025, I wrote that the firm had demonstrated its clear strength in enterprise AI, backed by its colossal cloud infrastructure, home-grown TPUs for AI training and inference, and first-hand access to the world’s data.

No doubt Alphabet will shout about this deal from the rafters at Google Cloud Next 2026.

Beyond the short-term victory laps, there are also significant long-term reputational gains for both Google and Apple that can arise from this deal.

After first announcing Apple Intelligence at WWDC 2024, Apple has had a rocky 18 months of repeated delays. Users who bought the iPhone 16 Pro on the understanding that they’d soon have a revamped Siri are still waiting for the promised updates.

In the meantime, the 16 Pro has been eclipsed by an entire generation of devices.

Apple has long practiced a ‘slow but effective’ approach to both its hardware and software, in which it shows up years late to technological trends but executes them to such a high standard that its user base can’t complain. Apple’s approach to AI has been no different, as we’ve previously noted.

With the iPhone, we’ve seen it with ultrawide lenses, high refresh rate panels, and later this year could see Apple do it again with a foldable.

But Apple users have no qualms about adopting competing software in the meantime – in the form of Gemini, OpenAI’s ChatGPT, and Anthropic’s Claude. Clearly, Apple has decided that its customers have waited too long and Google is the most reliable partner for immediate results.

Logan Kilpatrick, product lead for Google AI Studio and Gemini API, took to X in July to announce that the firm had processed 980 trillion tokens in May, across its products and APIs. This number has since risen to 1.3 quadrillion monthly tokens, with no notable outages across Google’s AI portfolio nor significant limitations imposed on free tier users.

Note also that this is a “multi-year” deal, with Google involved in crafting a range of new foundation models for Apple. Here the company can leverage its unparalleled access to data across Search, YouTube, Google Photos, and more, at a scale Apple can’t match.

The exact interdependencies between the Gemini family of models and Apple Intelligence is still to be determined, with Apple insisting the service will continue to operate both on device and using its Private Cloud Compute infrastructure.

In all likelihood, however, technical information will be kept at a minimum as Apple seeks to uphold a seamless AI experience for its iOS and macOS users.

FOLLOW US ON SOCIAL MEDIA

Apple, Google, and other browser makers have rolled out patches for zero-day bugs that are already being used by threat actors in "sophisticated" attacks.

Google noted that an exploit for one of the bugs exists in the wild already and was spotted by its Threat Analysis Group – that largely works on serious attacks led by state actors or similar – and was sorted via coordination with Apple engineers.

"For these Apple- and ANGLE-related issues, the quiet, coordinated disclosure strongly suggests the vendors viewed the bugs as high-risk and potentially already known to capable adversaries," Douglas McKee, director of vulnerability intelligence at Rapid7, told Dark Reading.

One of the bugs impacts other browser makers using Chromium, including Microsoft Edge and Vivaldi, which have also rolled out patches this week.

Patches for Apple

Alongside a set of other updates, Apple issued emergency patches for two issues in WebKit for devices running versions of its software before iOS 26. The flaws, CVE-2025-14174 and CVE-2025-43529, were credited, in full or in part, to Google Threat Analysis Group.

The first meant that accessing a webpage with "maliciously crafted" content could lead to arbitrary code execution. "A use-after-free issue was addressed with improved memory management," Apple said in a support document.

For the second, Apple said that malicious websites could lead to memory corruption, saying the issue was "addressed with improved validation."

"Apple is aware of a report that this issue may have been exploited in an extremely sophisticated attack against specific targeted individuals on versions of iOS before iOS 26," the company said in a support document detailing both issues.

The patches for the zero-day flaws are available for devices going back to iPhone 11, iPad Pro 12.9-inch 3rd generation, iPad Pro 11-inch 1st generation, iPad Air 3rd generation, iPad 8th generation, and iPad mini 5th generation.

Apple gave little extra detail about the zero-day flaws, saying it "doesn't disclose, discuss or confirm security issues". However, the tech giant issued a patch for the bugs alongside a set of other security issues that included further fixes for WebKit, the ScreenTime tool and more.

Google patches flaws

Google revealed its flaw via an update to the Stable Channel for the desktop version of its Chrome browser, crediting the discovery of CVE-2025-14174 to the Google Threat Analysis Group as well as Apple Security Engineering and Architecture.

Google had initially patched the flaw last week without any details beyond a "high" rating, but has now filled in a few details.

That flaw, labelled as a high-risk vulnerability, was first spotted December 5 and is caused by out-of-bounds memory access in Chrome's Almost Native Graphics Layer Engine (ANGLE), which is also used by WebKit, hence the impact on Apple.

"Out of bounds memory access in ANGLE in Google Chrome on Mac prior to 143.0.7499.110 allowed a remote attacker to perform out of bounds memory access via a crafted HTML page," said the CVE.org support page for the flaw.

"Google is aware that an exploit for CVE-2025-14174 exists in the wild," Google added in a blog post.

The company also patched two other medium-level flaws in the stable channel at the same time.

FOLLOW US ON SOCIAL MEDIA

When Apple first released a Pro model of iPad 10 years ago, the main aim was to provide a laptop-like experience in a tablet form. It wasn't necessarily a success on that front, but it proved popular regardless and gave the world's most popular tablet a premium tier.

The latest model, the M5 iPad Pro, is so good and intuitive that it could feasibly be a laptop replacement (depending on your job, of course). It features a new chip, an improved interface, and long battery life – something Apple seems to have mastered of late. It's for creators, it's for tablet-lovers, and yes, it is also for business uses.

Apple M5 iPad Pro review: Design

The design, like with all the MacBooks, hasn't seen much change. So, outwardly, the M5 iPad Pro is no different from the M4 (and older) models. The question is, would you want it any other way? The look of an Apple device is part of the appeal, and they're always easily distinguishable from Android tablets. In short, it's an iconic design, and it still, after all these years, looks great.

It only comes in either Silver or Space Black, which gives it a more professional (and slightly bland) look. And you can choose either a 13in or an 11in model. ITPro has the latter, which weighs just 444g. The tablet is very thin at just 5.3mm thick, though not as thin as the Samsung Galaxy Tab S11 Ultra (5.1mm). And despite that, it does still feel sturdy in the hand with its robust aluminum chassis.

Around the device, it has the usual curved edges, a minimal bezel encasing the display, buttons for power, volume up and down, and a Thunderbolt 3/USB 4 port at the bottom (if you hold it in portrait mode). You also have a 12MP webcam at the top of the screen (in landscape mode), which gives it that laptop feel. But that is mainly for iPad's Measure app and scanning capabilities.

Apple M5 iPad Pro review: Display

The iPad is all about the screen, and that is certainly the case with the iPad Pro. It features an 11-inch Ultra Retina Tandem OLED panel with impressive specs: 2420 x 1668 resolution, 264 ppi, and up to 120Hz of buttery smooth screen refresh. The whole thing is bright, vivid, and silky. What's more, you can pay a little more for the Nano Texture glass and do away with unsightly glare.

With our colorimeter, the Pro produced some impressive results, perhaps not as high as the Samsung Galaxy Tab S11 Ultra, but still of a high standard. For sRGB color gamut coverage, the Pro scored 97.3% and 100.2% for volume (99.7% for the Ultra).

The Pro is quite nice to use in low-light settings with its automatic controls for dark mode and nighttime settings. Writing up documents with the yellow-tinted 'night shift' setting certainly helps the eyes. But this, unmistakably a tablet for the more fun things in life; drawing, streaming, scrolling social media – all of which it excels at.

Apple M5 iPad Pro review: Keyboard and pen

One of the best additions to the iPad in recent years is the magnetic keyboard with its sturdy stand. We can't tell you how many tablets fall in this area – Huawei and OnePlus, we're looking at you – but Apple gets it spot on. The magnetic connection is easy to adjust and strong; you get two tilt angles, and it looks very neat and tidy. Add the all white Apple Pencil Pro on top, and it's very pretty.

The keys are also very good. Not MacBook Pro good, but also not far off. They have pretty backlighting, plenty of travel, and a nice muffled clack. Likewise, the trackpad is also very good. It's big enough and highly responsive.

The Pencil Pro is probably the overall star accessory and part of the reason the iPad has become such a stalwart in the art and illustration world. It is smooth and glides across the screen. The thickness of the pen itself is comfortable in the hand, similar to the OnePlus and Lenovo styluses. It doesn't replicate the feel of pen on paper, but instead makes digital drawing a worthy alternative.

Apple M5 iPad Pro review: Specs and Performance

Apple released a new iPad Pro alongside a new MacBook Pro and Vision Pro, but really, it was the launch of the M5 chip. For the iPad, it is either a 9-core or 10-core configuration of M5 and a 10-core GPU, and a 16-core Neural Engine. You can have up to 16GB RAM and as much as 2TB of storage (our review unit had 256GB).

Unsurprisingly, having the most powerful flagship processor in the iPad Pro results in the year's top tablet benchmark. In GeekBench 6, the M5 Pro hit 4193 for single-core and 16777 for multi-core workloads. Samsung's S11 Ultra managed 2,699 and 8,434 with its Mediatek 9400+ chip, while the OnePlus Pad 3 hit 3130 and 9083 with a Snapdragon 8 Elite processor.

There's also a big upgrade for AI capabilities on the iPad Pro; the GPU now has neural accelerators built into each of its ten cores. This is about boosting on-device AI processing speeds for features like the handwriting recognition or AI voice translations in Audio Logic Pro.

However, the most impressive upgrade, particularly for business use, is the battery. Our 11in M5 iPad Pro lasted an impressive 18hrs and 12mins in our looped video test. That was slightly longer than the S11 Ultra and about 2hours more than the OnePlus Pad 3.

The M5 iPad Pro supports iPadOS26, which brings a wealth of new tools and features. And that also includes the visual refresh of Liquid Glass, Apple's 'translucent' interface design.

Users will enjoy the new menu bar, which drops with a downward flick from the top of the screen. It works in the same way as it does in the macOS app – so potentially it can replace your laptop in this regard. The bar will soon be customizable so that it will become more important for work-related tasks, like it is for a MacBook. Apple Intelligence is also getting more useful with an integration into the Shortcuts app.

Apple M5 iPad Pro review: Is it worth it?

With the new M5 chip, Apple has finally turned its flagship tablet into a bona fide laptop replacement. The keyboard also does some of the work here, but the power, the speed, and the software tweaks all make for a brilliant device to use for work.

Could it be cheaper? Prices start at $999 (£999) for the 11in model, and that is less than the S11 Ultra, but significantly more than the OnePlus Pad 3. So it could be a little more budget-friendly, but it's better, in our opinion, than both those alternatives. This is, hands down, the most powerful tablet around.

Apple M5 iPad Pro specifications

Apple M5 is here, and surprisingly, unlike the multi-product release of the M4, it comes first in the MacBook Pro, iPad Pro, and Vision Pro. And, it is only the 14in MacBook Pro with the M5 chip, as the 16in model uses the M4 Pro or Max. And there is no M5 Pro or Max, as yet.

Now, there is a clear pattern over the last four years for the MacBook Pro, where very little changes in the aesthetic, and the main point of interest is the new silicon. This has roughly translated to a 20% power increase with each chip, more AI capabilities, and an ever-expanding battery life. And this is how the tech giant is pleasing workers, creatives, and Apple-aficionados.

You can have it now, in Silver or Space Black, starting at $1,599 (£1,599), with a dizzying array of configurations to choose from, including up to 4TB of storage. But does that make the M5 MacBook Pro enough of an upgrade?

Apple M5 MacBook Pro 14in review: Design

You'll have to take our word for it that the images in this review are of the M5 MacBook Pro, because it is nearly impossible to distinguish it from the M4 model (or the M3 one, for that matter). And that isn't necessarily a bad thing. It's a beautiful machine, made with recycled aluminum, and with the same old Apple logos and design features we know so well. It's a timeless design, and it needs to be, because it has been used time and again.

The 14in model comes with a selection of ports, which can seem like you have decent options, but ultimately that will be down to you and your specific needs. There are three Thunderbolt 4 slots (all of which take charge and connect to monitors), an HDMI 2.1 port, a full-sized SD card reader, and a headphone jack. And, a MagSafe 3 port with a USB-C connector.

The MagSafe charger creates an issue for people in Europe. Namely, you won't get a power brick in the box of your new MacBook Pro; you have to pay £59 extra for that. And the reason is the EU's Common Charger Directive, which will apply to laptops from 28 April 2026. For the last three years, the EU has required manufacturers to use USB-C connectors as standard, but has also decreed that chargers should not be bundled with devices in a bid to reduce e-waste. This has been the case for smartphones and other devices for a while now, but it will now include laptops. In true Apple fashion, its charger is actually a USB-C to MagSafe connection (meeting the EU halfway), but you can, of course, charge the MacBook Pro via one of its other USB-C ports.

Apple M5 MacBook Pro 14in review: Display, keyboard, and trackpad

The 14in MacBook Pro comes with the same display specifications as the 14in M4 model. A Liquid Retina XDR display, a 3024 x 1964 resolution, and up to 120Hz refresh rate with the ProMotion setting. The one noticeable change is that you can now boost the mini LED backlight, making it brighter. Apple notes that the panel can go up to 1,000 nits of brightness for SDR content. You can also pay extra for the Nano texture display, which is basically a layer that stops glare – and it does a pretty good job, in that regard.

Compared to the M4, there is a slight increase in quality. With our colormieter, the M5 MacBook Pro produced some pretty impressive scores for both gamut color coverage and volume – 99.7% and 144.1%, respectively. The M4 managed 96% and 98.9%. Its Adobe scores were also great (85.9%), as was its peak brightness (609.44cd/m2).

When it comes to the keyboard and trackpad, we literally could have copied and pasted from the 14in M4 review, but again, it's not necessarily a bad thing. The keys are nice and punchy, beautifully backlit, and with decent travel. The fingerprint reader in the top right is quick. And the trackpad is nice, big, and smooth. Plus, that haptic feedback is still so satisfying to use.

Apple M5 MacBook Pro 14in review: Specs and performance

With every iteration of the M-series, Apple has managed to provide significant performance increases across the board, taking each MacBook Pro a little bit further away from the competition. However, for the 14in M5 model, there are some quite impressive benchmarks for GPU performance. The reason is down to Neural Accelerators that have been added to each graphics core, a change that uses the technology from the AI-focused Neural Engine. By adding it to each of the 10 GPU cores, Apple has increased both the Pro's AI capabilities and its graphics performance.

In GeekBench 6, the M5 scored 4248 for single-core workloads and 17600 for multi-core ones. Both of which upped the 3899 and 15234 scores of the M4 MacBook Pro. However, what's more impressive is its GPU performance. We also ran the GPU benchmarks in GeekBench 6, and the M5 Pro came back with a whopping 48228. This graphical prowess was backed by its Cinebench 2024 scores, with GPU results for 1179 multithreaded performance and 201 for single.

There might be some question marks over who this new MacBook is for – the average office bod isn't going to be pushing the GPU that hard, and those who do that sort of work would arguably want the M5 Pro chip when it's available. But you're not going to complain about receiving the 14in M5 MacBook Pro for work. It is fast, it is cool, and crucially, it has unbelievably good battery life. In our looped video test, it lasted a whopping 20hrs and 31mins. And, like last year, it lasted longer than our actual looped video. The only laptop tested by ITPro to beat that is the 14in HP OmniBook 5, which recorded 20hrs, 40mins.

Apple M5 MacBook Pro 14in review: Features

As with all new laptops, AI is the word. Much of the M5 Pro's new capabilities stem from that super GPU and Neural Engine, but also the new macOS "Tahoe". The Sequoia operating system will be remembered for introducing Apple Intelligence, but it's Tahoe that looks to be the bigger beneficiary of the AI suite.

There are new on-device features, such as Live Translation in Messages, Phone, and FaceTime. Shortcuts are improved with new features, like automating actions where you can string them together in an automated workflow. And the whole thing just looks slightly different, and, in our opinion, better. The design change to the Liquid Glass improves elements like the menu bar or the app windows, making them more vibrant and appealing. Features like Continuity and Spotlight now offer new ways to work across devices. Apple has really nailed the way you can use both Mac and iPhone in tandem with the new Phone app for Mac.

If there is one feature that isn't changing, it's Wi-Fi 6. Why this is still on the MacBook Pro is strange, and even more so when you consider the M5 iPad Pro, which has Wi-Fi 7.

Apple M5 MacBook Pro 14in review: Is it worth it?

On battery life alone, the MacBook Pro is an office champion. But then, in that regard, so was the M4 model, and the HP OmniBook 5. What we're really looking at here is improvements in graphic performance and more AI capabilities. Which does beg the question, who is this 14in model for? Seems a bit too juiced for the average office bod. What's more, the M4 model of MacBook Air is significantly cheaper.

However, just because it is a bit too similar to the M4 model doesn't make it a bad laptop. It is the opposite, a stupendously great one, with a beautiful, bright display, brilliant graphical performance, and a ridiculously good battery.

Apple M5 MacBook Pro 14in specifications

The UK government has reportedly renewed its demand that Apple provide it with backdoor access to encrypted iCloud user data.

Earlier this year, it backed down on the issue after a furious response from Donald Trump. Apple, meanwhile, withdrew its most secure cloud storage service, iCloud Advanced Data Protection, from the UK in response.

Now, though, according to the Financial Times, the government is once again asking for access. This time, however, the technical capability notice (TCN) would apply only to the data of British citizens, rather than worldwide.

Both Apple and the Home Office are restricted by law from discussing the TCNs, which are issued under the UK Investigatory Powers Act.

Trump described the UK's earlier demands as 'something that you hear about with China' – and the Electronic Frontier Foundation (EFF) is inclined to agree, saying that the new TCN is just as dangerous as the first.

"As we've said time and time again, any backdoor built for the government puts everyone at greater risk of hacking, identity theft, and fraud," said EFF security and privacy activist Thorin Klosowski.

"It sets a dangerous precedent to demand similar data from other companies, and provides a runway for other authoritarian governments to issue comparable orders."

According to the FT, the US is no longer pushing the UK government to drop its demands. However, said Matthew Sinclair, UK senior director of the Computer and Communications Industry Association (CCIA), the government's decision to limit its request to UK data doesn't answer concerns.

"It is very concerning that reports again suggest that the Home Office is undermining security for British users of important cloud services," he said.

"Even if some changes have been made to reduce the extra-territorial impact, it is clear that any vulnerability will have implications beyond the UK's borders, creating a weak link in global efforts to improve security online. Changes to this law are urgently needed to stop this fundamental threat to encryption and ensure that the UK is not holding up important security improvements."

And, said Privacy International, end-to-end encryption can't be removed from iCloud services for the UK only when those services are used worldwide.

"Unless Apple plans to build walled-off systems just for UK users, so it can then undermine the security of UK users alone for the UK government, then this new order could have a quite similar impact to the old one," it said.

"This could also put at risk information like user passwords, location, and health data. By using a secret order to undermine the security of Apple products, the UK government is making security harder for us all."

And, it warned, "If this new order isn't stopped, the UK government will likely issue similar orders to other companies, too."

MORE FROM ITPRO

• I think the UK government's attempt to strong-arm Apple into giving it an ADP backdoor is a travesty – and so does most of the industry

Apple has issued an emergency security update for iPhone, iPad, and Mac users to patch a critical zero-day vulnerability that it said has already been exploited in the wild.

The out-of-bounds write vulnerability, designated CVE-2025-43300, was discovered by Apple security researchers and affects the Image I/O framework, which enables applications to read and write various different image file formats.

"Processing a malicious image file may result in memory corruption,” the tech giant said in an advisory.

An out-of-bounds write vulnerability occurs when attackers write data outside a program's allocated memory boundaries, potentially overwriting critical system information - in this case by sending a malicious image delivered via email, messaging apps or malicious websites.

If the victim opens the image, it could allow the attackers to carry out arbitrary code execution, potentially allowing them to take full control of an affected device and spy on users, steal data or install further malware.

The flaw affects the iPhone XS and later models, according to Apple. Devices impacted include:

• iPad Pro 13-inch
• iPad Pro 12.9-inch, 3rd generation and later
• iPad Pro 11-inch, 1st generation and later
• iPad Air, 3rd generation and later
• iPad 7th generation and later
• iPad mini, 5th generation and later

Worryingly, the flaw has already been exploited in the wild.

"Apple is aware of a report that this issue may have been exploited in an extremely sophisticated attack against specific targeted individuals," the company said - without giving any further details.

Apple urges users to update immediately

While this implies that the general public is unlikely to be targeted, Apple is advising users to install the new update, which includes improvements to bounds checking, immediately, via Settings, General and Software Update.

The chances are high that the successful attackers are nation-state actors. Similar exploits of Apple devices in the past have been linked to the use of spyware such as Israeli firm NSO Group's Pegasus software.

Already this year, Apple has patched five zero-days that have been exploited in the wild. These included:

• CVE-2025-24085
• CVE-2025-24200
• CVE-2025-24201
• CVE-2025-31200
• CVE-2025-31201

There were half a dozen more last year, figures show. The company also last month patched a Safari vulnerability - CVE-2025-6558 - residing in an open source component.

Google reported it had been exploited as a zero-day in the Chrome web browser.

MORE FROM ITPRO

• SharePoint flaw: Microsoft says hackers deploying ransomware
• Experts welcome UK government plans to back down in Apple encryption battle
• Zero-day exploits: How risky are they for businesses?

The UK government is reportedly planning to back down on plans to force Apple to share user data.

Tulsi Gabbard, US director of national intelligence, confirmed the move in a post on X, revealing that Downing Street will withdraw plans which would see Apple required to provide a “back door” to access user data.

“Over the past few months, I’ve been working closely with our partners in the UK, alongside @POTUS and @VP, to ensure Americans’ private data remains private and our Constitutional rights and civil liberties are protected,” Gabbard wrote.

“As a result, the UK has agreed to drop its mandate for Apple to provide a “back door” that would have enabled access to the protected encrypted data of American citizens and encroached on our civil liberties.”

The decision to back down follows a months-long war of words between Apple and the UK government over the tech giant’s Advanced Data Protection (ADP) tool, a feature that uses end-to-end encryption and allows users to protect personal data.

The feature even prevents Apple from accessing personal information.

In December last year, the government filed a format notice demanding that Apple grant authorities the right to access encrypted data. This was applicable to users globally, and not specifically in the UK.

The order sparked widespread criticism at the time from privacy experts and industry stakeholders. Legal experts warned that the move could harm industry perception of the UK, noting that the country would “no longer be seen as a safe destination for personal data”.

In response, Apple refused to comply and announced plans to remove access to ADP for UK users. The tech giant began legal proceedings to challenge the demand.

“As we have said many times before, we have never built a backdoor or master key to any of our products or services and we never will,” the company said in a statement at the time.

Apple u-turn “hugely significant” for UK government

The UK government is yet to confirm the decision to back down, however, the rumored decision has been welcomed by security experts.

Charlotte Wilson, head of enterprise at Check Point Software, said the decision to withdraw the demand is “hugely significant”.

“Once you create a master key to encrypted data, it is not just governments that can use it,” she said. “Criminal groups and hostile states will try to exploit it too.”

Wilson added that breaking encryption “not only weakens privacy, but also weakens security for everyone.”

The Apple order was issued under the Investigatory Powers Act, which means orders issued to other companies are essentially shrouded in secrecy.

Wilson said there are still concerns that these powers “remain on the statute books”, meaning that future governments may attempt a similar move in the future.

“Even if this particular demand has been withdrawn, the Investigatory Powers Act still allows future governments to try again,” she said.

“That leaves citizens, campaigners, and businesses in a state of uncertainty about whether the tools that protect their data could be undermined at any point.”

MORE FROM ITPRO

• NCSC expert urges businesses to follow geopolitics as defensive strategy
• The encryption stand-off is getting weirder
• What is end-to-end encryption and why is everyone fighting over it?

Mini PCs are all the rage right now and have been for at least two years. The Apple Mac Mini M4 has taken the 'mini' part and truly run with it, for it is a tiny cube of a computer. Don't let that fool you, though, for it's a powerful little box with an M4 chip.

Such is the appeal of the mini PC, Apple has not only popped an M4 chip in (M4 Pro in higher spec models), it has also redesigned the mini, making it smaller and changing up the port array.

Before we get carried away, there is no bundle discount with Apple; it's best to look to channel partners and resellers for discounts. The company is looking at small businesses here, which are more likely to have all the peripherals already.

Apple Mac Mini M4: Design

There is very little to say about the design of the M4 Mac Mini that doesn't leave you feeling like you're describing a kitten or a novelty baby outfit, because it is the cutest little computer we've ever seen. A truly mini machine that fits so snugly in your hands that you endup cradling it. And with that minimalist design that runs through the whole Apple hardware range, it is just a lovely addition to your workspace.

Neatness is its biggest appeal; it's small and takes up next to no desk space, but also doesn't really leave anything out (aside from perhaps a USB-A port). You have five USB-C slots (two front, three rear), an HDMI port, a headphone jack, an ethernet slot, and the power socket. What's more, everything is set in a way to keep things neat – all the ports are lined up so that everything fits in, nothing has reason to be tangled. From the power socket to the USB-C slots, there is ample room, with no overlap. There's only really a strange placement of the power button to complain about, but that isn't strange to people who use Mini PCs for 1U racks.

Apple Mac Mini M4: Specs and performance

Our Mac Mini review unit houses an Apple M4 chip with a 10-core CPU, a 10-core GPU, and a 16-core neural engine. We also have 16GB of unified memory and 256GB of SSD – this is the base model of Mac Mini (£599-$499), and you can go up to 64GB RAM, 8TB SSD, and the M4 Pro chip but you'll be paying out well over a grand (£1,399-$1,299).

In Geekbench 6, the M4 chip shows consistent scoring across the board, with the Mac Mini in the same range as the MacBook Pro and the MacBook Air – it's also similar to the M4 Pro chip in the more expensive Mac Mini, though the multi-core score is noticeably higher. For the base M4 model, we noted a single-core score of 3,496 and a multi-core score of 14,505.

The latest version of macOS (Sequoia) is as smooth as ever, even with intensive photo editing or some light gaming – the mini is also quiet, with virtually no sounds of strain, no matter the task.

Perhaps the best part of the M4 chip, however, is the power efficiency. We've already seen this in the MacBook Pro, where there is a clearer benefit with its optimized battery performance. For the Mini, it is less obvious, but it does run cooler (part of the reason it's so small).

Apple Mac Mini M4: Features and peripherals

Apple Intelligence is also present on the Mac Mini. As with the MacBook Pro and Air, it is good, though still in its infancy. It's a similar experience to Microsoft's Copilot, with text generation, image editing, and search optimization. Having Siri answer your questions is helpful, as is the email summarizations.

The new mini only comes with the device itself and the power cord. If you want a keyboard or mouse, or trackpad, you need to purchase those separately. The thinking here is that consumers will likely already have them with older models of Mac Mini, and businesses usually have monitors and so on that are already set up on desks. ITPro, for example, still has the Apple Studio Display, the Magic Mouse, and the keyboard from a previous Mac Mini review.

If, however, you do need to purchase peripherals, a new Magic Keyboard will set you back $99(£99), a Magic Trackpad $129 (£129), and the Magic Mouse is $79 (£79). All of which have slight design changes with universal USB-C slots and cables, rather than the old Lightning cable – which is great. Although the charging port on the mouse is still on the underside of it, which is somewhat odd, it's problematic as you can't use it while it's charging.

Apple Mac Mini M4: Is it worth it?

One other area that Apple has made changes in is its pricing. While on a small drop compared to previous models, the $599 (£599) starting price is extremely competitive – noticeably cheaper than our last reviews from Geekom and Beelink.

The lack of peripherals in the box may be off-putting to some, but from a business standpoint, it isn't much more expensive to add the keyboard and mouse; most other vendors do the same, and office desks are generally already kitted out with monitors.

The M4 chip is super powerful, the user experience is great, and it fits on your desk so well you forget it's there. That's how all technology should be, useful and unobtrusive. We love the Mac Mini; it's tiny, tidy, and terrific.

Apple Mac Mini M4 specifications

At $899, the Apple MacBook Air 15in M4 is surprisingly affordable. Surprisingly, because Apple isn't really known for that, and MacBooks tend to be expensive purchases. However, with the looming switch off of Windows 10, Apple senses an opportunity to capture some new business – specifically, small businesses – and now might be the best time to invest in its latest laptops.

There's little change to the design, save for a fancy new blue color option, and most of the upgrade centers around that super M4 chip. Compared to the M4 Pro, there are a few drops in quality, and no Wi-Fi 7. But you can have the M4 MacBook Air with 16GB RAM and 256GB of SSD – or configure them up to 32GB and 2TB, respectively.

Apple MacBook Air 15in M4: Design

So the new color for the MacBook M4 is a very light, almost grey, sky blue. A very particular sky blue, like the lightest sky you have ever seen. And in truth, there are times it just looks grey. But in the right light, it's sky blue and we like it. We're talking about the color a lot because that is really the only change in design compared to the M2 and M3 models of MacBook Air.

In a way, this is good, as the M3 was beautiful, as most MacBooks are. The M4 Air is pretty in a minimal way, with little noise around the chassis and keyboard, and just that webcam notch the only blemish. The build quality is cool and ridged – there is absolutely no flex in the lid. It is also not very light at 1.5kg, not very Air-like, but then again it isn't exactly heavy; the 15in size is harder to get in a backpack, but you'll not notice it when you have slipped it in.

Apple M4 MacBook Air: Display

The 15in Liquid Retina display on the MacBook Air is bursting with quality. It has a 2,880 x 1,864 resolution, covers 100% of the DCI P3 color gamut in its default mode, and reaches over 500cd/m2 for brightness (512cm/d2 in our tests).

Color accuracy was also high; with our colormeteier, and measuring in the default color mode, we recorded the MacBook Air showcasing 98.7% for sRGB gamut color coverage. The scores were largely similar to the MacBook Pro, which is impressive. In truth, video quality is not as good as your best OLED screens – you don't quite get those inky blacks – but it is nonetheless a high-quality screen.

And yet, there are still complaints. Apple will never go for a touchscreen on a MacBook, not a massive deal, to be honest, but they can and should be upping the refresh rates. 60Hz on a screen in this day and age is criminal. This, a MacBook Air, with all its quality elsewhere, should be giving us at least 90Hz for smoother scrolling. We found this an annoyance going up and down spreadsheets and documents, where everything blurs if you go too fast.

Apple MacBook Air 15in M4: Keyboard and trackpad

Apple has made major changes to its keyboard since its butterfly-switch catastrophe a few years ago. So the keys on the M4 MacBook Air are largely the same as the ones on the M3 and M2 models – and that's a good thing really. All three models have great keyboards, so there's little complaint here at ITPro.

The trackpad is largely the same as on the Pro, in that it is rather big and also very good. Single finger taps, multi-touch gestures, and a refined forced feedback engine that lets you double-click and prod with ease. The whole experience is effortless and makes navigation simple.

Apple MacBook Air 15in M4: Specs and performance

As with the Pro, the most noticeable upgrade on the MacBook Air is the M4 chip. The CPU count is up from eight to ten compared to the M3, and there are four performance cores and six efficiency cores. The GPU core count is the same a the M3, but there is a performance improvement.

In Geekbench 6, the M4 Air produced stellar results, with 3,762 for single-core performance and 14,872 for multi-threaded performance. These are only slightly below the scores of the MacBook Pro, but way ahead of the M3 MacBook Air. Apple is far and away from other vendors here, with only the Microsoft Surface Laptop 7 coming close to its multi-threaded performance.

When it comes to battery life, the M4 MacBook Air is better than the M3, but not by much. However, it isn't as good as the M4 Pro by a significant distance. In our looped video test, the MacBook Air lasted 15hrs and 39mins – almost five hours off the MacBook Pro.

Apple MacBook Air 15in M4: Features

The M4 MacBook Air supports a range of features, and we will start with the standard stuff. Only Wi‑Fi 6E (not 7 for reasons unknown), and Bluetooth 5.3. There's a MagSafe 3 charging port and 3.5mm headphone jack. You get two Thunderbolt 4 (USB-C) ports with support for charging, DisplayPort, and Thunderbolt 4 (up to 40Gb/s) USB 4 (up to 40Gb/s).

Getting to the more advanced stuff, you have Apple Intelligence, the company's AI engine. On a laptop, as we found with the Pro, there are some handy features to enjoy here – though nothing much different from your average copilot. The writing tools feel like the best application (so far) as they are genuinely useful, helping with proofreading and summarisation. The former is on par with tools like Grammarly, while the latter is slightly less annoying than Google's Gemini.

Siri has also had a brain boost, as it seems a little bit cleverer; you can use a quick type search bar in the upper right of the display and learn new things quicker than clicking and typing in Google.

Ever more useful is the fact that you can now connect up to two external 6K monitors while continuing to use the MacBook's display – if you recall, the M3 model could only do so if you closed the lid. So that's a real improvement.

Apple MacBook Air 15in M4: Is it worth it?

The first question to ask yourself (or your business) here is what do you want the Air for? If you're a new business with a fondness for Apple hardware, the Pro, though more expensive, is the better laptop (longer battery life, Wi-Fi 7, slightly faster). And it feels more like a work machine, too.

That being said, the Apple MacBook Air M4, with its larger 15in display, is a smashing machine. It looks good (the blue is swish), it has a nice keyboard, it's fast, and the battery life is decent. Prices start at $899 (£899), which is surprisingly affordable.

Apple MacBook Air 15in M4 specifications

The Apple MacBook Pro 14in M4 has a lot more riding on it this year. With the looming end to Windows 10 support fast approaching, a mass hardware refresh is expected. The system requirements for Windows 11 call for more powerful laptops. Here, Apple senses an opportunity; a chance to capture more users, particularly those at smaller and newer businesses.

You can get the latest MacBook for $1,599 (£1,599) with the basic configuration (16GB RAM and 512GB storage). Or you can pay $133.25 per month for 12 months – business customers will likely enjoy a discount for bulk purchases.

Apple MacBook Pro 14in M4: Design

There is a level of tidiness to Apple's hardware that few vendors can match. It's what makes its iMac screens popular for front-of-house use, or its Mac Mini machines adorable and minimalist. And its MacBook designs have influenced other vendors for almost two decades.

As such, nothing much has changed on the design compared to the M3 MacBook, though everything is neat and elegant with a smooth satin finish. The latest MacBook Pro is made from 100% recycled aluminum, with 90% recycled steel in the keyboard feature plate and 35% (or more) recycled plastic in 15 components, according to Apple. It comes in the traditional silver or (like our review unit) the more understated 'Space Black'. It looks and feels like a serious business machine, while still holding great appeal for consumers.

At 1.6kg the Pro is fairly heavy; it's a sizable hunk of laptop, and robust enough to give to employees without concern. There's a sturdy feel to the chassis and next to no flex in the lid. All in all, a solid machine that can handle the rough and tumble of remote working.

Apple MacBook Pro 14in M4: Display

As ever, Apple's Liquid Retina XDR display is on the MacBook Pro, and it's superb. Our review unit is the 14.2in model, with a 3024 x 1964 resolution and a 'ProMotion' 120Hz refresh rate that's buttery smooth. You can get a 16in model for $1,000 more, but the 14in is a good size for most office workers.

We measured the Pro's ability to reproduce colors in the sRGB color space, using the laptop's default color mode, and were left largely impressed. It hit 96% for color coverage and 98.9% for volume, with 70.1% for DCI P3. All very good scores (anything over 94% is excellent), matching the M3 MacBook Pro.

The panel can get extremely bright, too. Apple suggests it can sustain a brightness level of 1,000 cd.m2 with regular SDR content, though we only measured its peak brightness as 636.83 cd/m2 in the default setting.

Apple MacBook Pro 14in M4: Keyboard and trackpad

Typing on the MacBook Pro M4 is a dream with an excellent keyboard. Travel is a little on the shallow side, but the individual key action is great with nice positive click feedback and very little bounce or flex in the base. Once you get up to a decent typing speed, you'll notice a lovely sound – almost comforting – that the keys emit. Like a muffled drum. Pure bliss.

Its trackpad, too, is rather pleasurable. It's big, which is always useful, and highly responsive, also useful. We particularly like the single finger taps and the multi-touch gestures. However, the most lovely thing about it is the refined force feedback engine, which works so effortlessly, letting you click or double click with an easy prod in the tip left corner.

Apple MacBook Pro 14in M4: Specs and performance

When you look at what changes year on year with the MacBook Pro (and Air, or iPhone, or iPad, and so on), the Apple chip often shows the biggest improvement. And the M4 processor is another industry-leading laptop chip; it comes with a 10-core CPU, 10-core GPU, and a 16-core neural engine. There is also the normal 16GB of RAM and 1TB of SSD (other configurations are available).

It feels powerful; there is no amount of tabs or applications we could run that would truly test its limits. However, in Geekbench 6, the M4 Pro scored 3,899 for single-core workloads and 15,234 for multi-core. Strangely, that is a significant increase for single cores and a minor drop for multi-threaded versus the M3 MacBook Pro (3,184 and 15,704, respectively). But, still, vastly more than anything else on the market.

For our looped video test, we set the display brightness to 170 cd/m2, turn the machine on airplane mode, and run a 20hr looped video until the battery runs out. Astonishingly, the MacBook Pro M4 outlasted the 20-hour video with 4% left on its battery – there is no other laptop, tested by ITPro this year, last year, or seemingly ever, that has that level of staying power.

And, as with all PCs now, we must talk about AI for productivity. On the Pro, we have Apple Intelligence, which can do most of the stuff your Copilot+ PCs can, but with a few Apple additions and tweaks. You can use it to customize your desktop, add apps you the top right of the screen. Summarize and transcribe emails. And also, if you have an iPhone to match, you can scan documents onto the MacBook Pro, which can then be automatically read and edited with the text translator. Nothing groundbreaking, but we did find it very useful.

Apple MacBook Pro 14in M4: Is it worth it?

Whatever configuration you get, no matter the price, the MacBook Pro 14in M4 is worth it. As a consumer device, it is amazing. As a work laptop, it is phenomenal – a surefire way to please your employees. If you have a BYOD strategy and those devices are iPhones, then it is even more appealing. Those with M3 models might have reason to stay put, but anyone with an M1 (or older) will be getting a vastly more powerful machine and a battery life you'll scarcely believe.

Apple MacBook Pro 14in M4 specifications

Apple is pushing back against new app store rules imposed by the European Commission (EC), claiming the EU body has taken changes and a €500 million (£430m) fine too far.

The EC levied the fine in April, arguing Apple was in breach of anti-steering obligations in the EU Digital Markets Act (DMA). That refers to Apple's long-running ban on alternative app stores and rules that stop developers from telling users about discounts or other offers outside of the Apple platform.

Apple previously took a cut of up to 30% on sales inside apps and has been locked in a war of words over the practice with developers.

The tech giant complied with the app store rule changes last month in order to avoid additional fines that could reach 5% of its average revenue globally.

That included a wider range of commission rates up to 13%, alongside user acquisition fees, with the costs dependent on what marketing support the developer wants in the App Store as well as technical aspects such as automatic updates.

Developers will also be able to promote alternative ways of purchasing services outside the app, though Apple would charge a commission on sales advertised via apps in its App Store.

Apple is baffled, the EC is standing firm

Key to Apple’s argument, and appeal, is that the company claims this new system is "confusing" to business users.

"Today we filed our appeal because we believe the European Commission’s decision – and their unprecedented fine – go far beyond what the law requires," said Apple in a statement.

"As our appeal will show, the EC is mandating how we run our store and forcing business terms which are confusing for developers and bad for users."

Apple has yet to reply to a request for comment from ITPro.

The Commission, meanwhile, is standing by its decision. In a statement responding to the appeal, it said: "We stand ready to defend our decisions in court.”

The EC previously said that Apple shouldn't charge developers for the right to "steer" customers, suggesting that a commission on sales promoted in-app wouldn't meet requirements.

Political battle

A legal expert told The Guardian that Apple may be employing delay tactics.

"The blunt truth is that it is worth spending a few million on legal fees in order to disrupt and delay the development of a more open app ecosystem, which is a market that is worth many billions a year to Apple," Tom Smith, a competition lawyer at Geradin Partners, told the publication.

The app store quarrel isn't the only ongoing battle between the EU and tech companies. In April, Meta was hit with a €200m fine for charging users to get rid of ads, while last year Apple was fined €1.8bn over music streaming competition.

The appeal comes ahead of US President Donald Trump's 9 July deadline to settle a trade deal with the EU or risk 50% tariffs. The Trump administration has repeatedly said the EU is targeting US tech companies in retaliation for trade disruption and other political concerns, with Trump trade advisor Peter Navarro calling it "lawfare".

EU action against tech giants stretches back well before even the first Trump administration, with a €497 million fine against Microsoft in 2004 over Windows Media Player bundling that was later increased to €899 million for non-compliance.

That was followed by cases against Intel in 2009, an investigation against Google on search that began in 2010, and investigations into taxes that pulled in Apple and Amazon beginning in 2014.

That said, things have accelerated since 2017 with a €2.42bn fine against Google over price comparison shopping services. In subsequent years, EU lawmakers imposed a €4.34bn for Android and €1.49bn over AdSense, with other cases targeting Apple, Amazon and Meta in recent years.

MORE FROM ITPRO

• Apple throws cold water on the potential of AI reasoning
• Apple is offering rewards of up to $1 million to find critical flaws in its private AI cloud systems
• Apple: Our AI data was gathered responsibly

A report from Apple has cast significant doubts on the efficacy of reasoning models, going as far as to suggest that when a problem is too complex, they simply give up.

AI reasoning models have emerged in the past year as a beacon of hope for large language models (LLMs), with AI developers such as OpenAI, Google, and Anthropic selling them as the go-to solution for solving the most complex business problems.

However, a new research paper by Apple has cast significant doubts on the efficacy of reasoning models, going as far as to suggest that when a problem is too complex, they simply give up. What's going on here? And does it mean reasoning models are fundamentally flawed?

In this episode, Rory Bathgate speaks to ITPro's news and analysis editor Ross Kelly to explain some of the report's key findings and what it means for the future of AI development.

Highlights

"It sounds a bit obvious, but at the same time, the way that this has been, framed by a lot of providers, makes it seem like they're the be-all and end-all. But once you reach a level of complexity, yeah, you start to encounter some serious problems."

"What's interesting for me, right off the bat, is, what Apple is showing is that that's actually the opposite of what these are good at, that they are maybe more performant, maybe they're slightly better at solving complex problems up to a point, but that if you actually try and give them really complex problems, they are completely ineffective."

"Apple is claiming that OpenAI hasn't made a model that can think, to which my reaction was, did you think OpenAI had made a model that can think? "

• Subscribe to The IT Pro Podcast on Apple Podcasts
• Subscribe to The IT Pro Podcast on Spotify
• Subscribe to the IT Pro newsletter
• Join us on LinkedIn

Apple has suggested that AI reasoning models have clear limits when it comes to solving complex problems, undermining developer arguments that they are useful for tasks that a human would traditionally solve.

Reasoning models can solve more complex problems than standard large language models (LLMs) by breaking them down into a series of smaller problems which are solved one by one.

A host of major providers, including OpenAI, Anthropic, and Google have highlighted the benefits of reasoning models over the last year, touting these as a key weapon in the enterprise AI arsenal.

The paper, titled The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity, specifically names OpenAI’s o1 and o3 models, DeepSeek R1, Anthropic’s Claude 3.7 Sonnet, and the latest version of Google’s Gemini.

It found that while these models efficiently handled low complexity tasks, using relatively few tokens to solve the given problem, beyond a certain level of complexity they began to take much longer to respond, waste tokens, and return incorrect answers.

Apple’s research team stated that the benchmarks currently used to evaluate what it calls large reasoning models (LRMs) are flawed, as they tend to center around how well a given model can solve coding and mathematical problems.

The researchers argued the results of these benchmarks are susceptible to data contamination – in which the answers are accidentally incorporated into a model’s training phase – and don’t give researchers proper control over the variables.

In response, they devised new puzzles for the RLMs with an emphasis on logical reasoning and no requirement for the model to use outside knowledge to reach an answer. The results showed that beyond a certain level of complexity, the models demonstrated total failure.

“Through extensive experimentation across diverse puzzles, we show that frontier LRMs face a complete accuracy collapse beyond certain complexities,” the researchers wrote.

One of the puzzles researchers used was the Tower of Hanoi, in which disks are stacked in order of size atop a rod next to two others. The player must move the ‘tower’ of disks in from the first rod to the third by moving only one disk per turn, and without a larger disk ever being set atop one smaller than itself.

With just three disks at the start, the models were able to solve the problem and were graded highly for each step they took toward the solution. But beyond a set number of disks, the accuracy of all models fell to zero.

Researchers noted this was the case even when the solution algorithm was provided in the prompt, implying the limitation is inherent to reasoning models. Even with less complex problems, models were also found to ‘overthink’ – finding the solution early on, then wasting time and tokens considering other, incorrect paths to the solution.

Overall, the paper suggests that LRMs are far less generalizable than providers have suggested, with a tendency to reduce computation as problems increase in difficulty rather than increase it and power through to the solution.

The work was carried out by Apple researchers Parshin Shojaee, Iman Mirzadeh, Keivan Alizadeh, Maxwell Horton, Samy Bengio, and Mehrdad Farajtabar. Between them, the team includes experts in efficient LLM usage, AI and machine learning, as well as an intern focused on deep learning and reinforcement learning.

A huge blow for the ‘reasoning’ crowd

Apple’s research paper seriously rains on the parade of the world’s most prominent AI developers, most of whom have spent the past nine months shouting from the rooftops about the potential for reasoning models.

The exact claims about what ‘reasoning’ constitutes have always been hazy, to say the least. In its announcement post for o1-preview, OpenAI claimed it had developed a new family of models that could “spend more time thinking through problems before they respond, much like a person would” and even “refine their thinking process” based on mistakes.

Similarly, Google has hyped up how Gemini 2.5’s reasoning helps it handle more complicated problems while Anthropic even claimed that 3.7 Sonnet was specifically designed with “real-world tasks that better reflect how businesses actually use LLMs” rather than math and computer science problems.

This research paper throws these claims into serious doubt, particularly in its systematic approach to showing that LRMs get less accurate, not more, when presented with harder tasks.

Earlier LLMs have tended to prioritize information given at the beginning and end of large prompts and brush over the context provided in the middle. This has meant they’ve historically struggled with so-called ‘needle in a haystack’ problems, in which they are tested on their ability to respond to information buried somewhere in the middle of a prompt in excess of 128,000 tokens.

What we’re seeing here could be the reasoning model equivalent, a fatal flaw at the heart of the algorithms powering LRMs that cause them to lose ‘focus’ on a problem and inevitably fail when allowed to work on a problem for too long.

It’s worth noting that the researchers allowed the models to use their full ‘budget’ for thinking – for example, 64,000 tokens for Claude 3.7 Sonnet – but that beyond a certain level of complexity, the models simply stopped spending tokens on reasoning through the problem further.

No one wants AI models that give up if they decide a problem is too hard, not least if they’ve been promised these are the ones built for the hardest problems.

If the limitations are indeed inherent to our current training approaches, developers are facing the headache of quickly implementing new approaches and architectures that solve this issue.

For now, customers may start to question if the ‘thinking’ models they’ve been sold are truly up to the task.

MORE FROM ITPRO

• Apple: Our AI data was gathered responsibly
• How effective are AI agents?
• The risks of open source AI models

The European Commission has issued its first penalties under the EU Digital Markets Act (DMA), fining Apple €500 million and Meta €200m and requiring both to make changes to how they operate.

The DMA came into effect in 2023 and was created to regulate competition in the digital economy, in part by reining in abuse of power by larger platforms.

Apple was found to be in breach of "anti-steering obligations" over restrictions that ban apps from informing customers of offers outside the App Store, which the EC said limited customer choice.

The Commission has ordered Apple to remove that restriction, and the tech giant plans to appeal the fine.

Meta, meanwhile, was fined for its introduction of a model that would force users to pay for an ad-free service or consent to their data being used for advertising.

The Commission said that fell foul of an obligation under the DMA to give consumers the choice of a service that uses less of their personal data.

Apple and Meta respond

Both companies made it clear they disagreed with the fines. In a statement, Apple hit out at the Commission, suggesting the penalties show it is “unfairly targeting” the company.

The tech giant said the decisions are “bad for the privacy and security of our users, bad for products, and force us to give away our technology for free”.

"We have spent hundreds of thousands of engineering hours and made dozens of changes to comply with this law, none of which our users have asked for,” the statement added.

“Despite countless meetings, the Commission continues to move the goal posts every step of the way. We will appeal and continue engaging with the Commission in service of our European customers.”

Meta offered harsher criticism in its response, accusing the EC of treating US businesses unfairly.

"The European Commission is attempting to handicap successful American businesses while allowing Chinese and European companies to operate under different standards," Meta's Chief Global Affairs Officer, Joel Kaplan, said in a statement.

"This isn’t just about a fine; the Commission forcing us to change our business model effectively imposes a multi-billion-dollar tariff on Meta while requiring us to offer an inferior service," he added. "And by unfairly restricting personalized advertising the European Commission is also hurting European businesses and economies."

A ‘strong and clear’ message from the EU

Despite the complaints, Teresa Ribera, Executive Vice-President for a Clean, Just and Competitive Transition, said in a statement the fines send a "strong and clear message" to digital companies, adding that the legislation is a crucial tool to ensuring fair markets.

"Apple and Meta have fallen short of compliance with the DMA by implementing measures that reinforce the dependence of business users and consumers on their platforms," Ribera said. "As a result, we have taken firm but balanced enforcement action against both companies, based on clear and predictable rules."

"All companies operating in the EU must follow our laws and respect European values,” she added.

That reference to European values reflects geopolitical turmoil between the US and the rest of the world, according to Joe Jones, director of research and insights at IAPP.

"The fines land at a time of heightened scrutiny by the current US Administration on the application of EU laws to US companies," Jones said.

"The EU Digital Markets Act was even name-checked in an Executive Order issued by President Trump last February as facing scrutiny as part of the Administration’s work to “defend American companies and innovators from overseas extortion,” he added.

"Open questions include not only how will addressed companies respond to EU regulatory enforcement but how will overseas governments, including and especially the US, respond. The U.S. Administration has declared it will consider responsive actions like tariffs to combat certain foreign government policies levied against US companies."

Beyond the fines

As Meta noted, there's more to the decision than fines. Indeed, despite the seemingly large figures, the fines were much smaller than the EU could have levied, as high as 10% of their global annual turnover. Last year, Meta's earnings topped $165 billion and Apple's was $391 billion.

In a statement, the EC said that developers using Apple's App Store should be able to tell customers about alternative offers outside that store, such as discounted subscription offers. The EC added that Apple had failed to give a reason why such restrictions are "necessary and proportionate."

Beyond the fine, the EC has ordered Apple to remove such restrictions and not introduce similar restrictions in the future.

The EC also closed an investigation into Apple's tactic of not allowing its own pre-installed apps to be removed from its devices, such as its Safari web browser.

Following a "constructive dialogue", Apple made it easier to select a new default browser on iPhones, uninstall several Apple pre-installed apps, and change default settings for other tools such as keyboards and translation.

Alongside those two investigations, the EC also issued preliminary findings on Apple banning third-party app stores and requiring users to download apps via the official App Store.

The EC said that its preliminary view is that Apple isn't complying with the DMA by disincentivising developers from offering apps elsewhere via a new fee, the Core Technology fee, while also introducing strict eligibility requirements.

Apple can now respond to those findings ahead of a final ruling.

For Meta, it has been fined under DMA rules that mean "gatekeepers" must get user consent for combining personal data between services, and still offer an equivalent alternative service for any who refuse to consent.

This decision refers to Meta's "consent or pay" advertising model that was introduced in November 2023, which gave Facebook and Instagram users the option of consenting to their data being combined for personalized advertising or paying a subscription to opt out and receive an ad-free service.

The EC said that model didn't offer enough of an alternative service for those who opted out. A year later, Meta offered a new version that included a tier that used less personal data to display ads.

This particular fine applies only to the period up to November 2024 when those changes came into effect, as the regulator is still considering the changes.

Alongside the advertising data ruling, the EC also announced that Facebook Marketplace was small enough that it should no longer be designated under the DMA.

MORE FROM ITPRO

• How the EU AI Act compares to other international regulatory approaches
• DORA and why resilience (once again) matters to the board
• The fractured regulatory landscape tech companies face in 2025

Apple has announced plans to invest more than $500 billion in the US over the next four years, marking the tech giant’s largest-ever spending commitment.

The funding will go on a broad mix of projects, including expanding the staff and facilities in Michigan, Texas, California, Arizona, Nevada, Iowa, Oregon, North Carolina, and Washington, as well as on a new factory in Texas.

Apple said it will open the new advanced manufacturing facility in Houston to produce servers that support Apple Intelligence. The 250,000-square-foot server manufacturing facility, set to open next year, is expected to create thousands of jobs.

It will also double its US Advanced Manufacturing Fund from $5 billion to $10 billion, with a particular focus on promoting advanced manufacturing and skills development across the US.

Similarly, a new manufacturing Academy in Detroit will link Apple engineers and experts from universities including Michigan State with small- and medium-sized businesses, advising them on implementing AI and smart manufacturing techniques.

"We are bullish on the future of American innovation, and we’re proud to build on our long-standing U.S. investments with this $500 billion commitment to our country’s future," said Apple CEO Tim Cook.

As part of the investment scheme, Apple also plans to expand its research and development funding in areas such as silicon engineering, software development, and AI and machine learning, hiring around 20,000 people.

"From doubling our Advanced Manufacturing Fund, to building advanced technology in Texas, we’re thrilled to expand our support for American manufacturing," said Cook.

"And we’ll keep working with people and companies across this country to help write an extraordinary new chapter in the history of American innovation."

Apple’s big domestic push – but is it new?

Apple’s investment is all about increasing the amount of its domestic manufacturing, talent, and AI infrastructure. But how much of this is new? No one is sure.

Apple could be doing some creative headline writing here, given it boasted about a $350 billion investment in 2018 and a $430 billion investment in 2021, across overlapping five year timeframes.

The full sum also includes Apple TV spending – which had totaled $20 billion by 2024 per Bloomberg – and supplier spending.

That said, it’s clear that there’s a significant (if unspecific) figure being invested here in a strategic move to secure its supply chain. Since 2020, Apple has been sprinting to produce its own components.

It launched its first M series chips in November of that year, first for MacBooks and eventually across its full lineup of devices, and by 2025 has entirely replaced Intel chips with its in-house hardware.

Meeting this goal has meant heavy reliance on TSMC, particularly its many fabrication plants in Taiwan, where TSMC produces leading hardware including Nvidia’s most advanced GPUs for AI in data centers.

But Taiwan’s precarious geopolitical position with relation to China has for years cast doubt over whether this supply chain is viable in the long term.

The tech giant’s Texas server factory will be needed as Apple looks to expand its Apple Intelligence generative AI feature across its software offerings, given the company’s focus on manufacturing and operating as many of its products and services as possible in-house.

While Apple inked a deal with OpenAI last year, Apple Intelligence relies on in-house LLMs trained by Apple using its own servers and chips. As it looks to improve the sophistication of its on-device AI models, as well as offer users more advanced capabilities via its Private Cloud Compute, it will need to rapidly expand servers for training and inference.

Apple cannot replace TSMC, nor is it seeking to. Its AI servers currently use Apple silicon and the expanded servers made in Texas will run next generation M series chips produced by TSMC, according to a post on X by Ming Kuo, securities analyst at TF International.

But in slowly expanding its domestic manufacturing capabilities, Apple is putting itself in a far more stable supply chain position.

Its commitment to its Detroit manufacturing academy and beefed up research and development funding for silicon engineering and AI could also prove transformative in the long-run as it seeks to empower US workers with the necessary skills to meet chip demand.

TSMC has faced union backlash over the large number of workers from Taiwan employed at its Arizona facility, as reported by Tom’s Guide, with around half of all jobs created at the site so far having gone to Taiwanese workers.

MORE FROM ITPRO

• Investment in UK's AI sector surged last year
• Stargate project: OpenAI, Oracle pledge support for $500 billion AI infrastructure drive
• Cloud spending projected to grow 19% this year on back of strong 2024

Earlier this month reports emerged that the UK government issued an order to Apple to give it access to users’ encrypted data, reigniting the debate around digital privacy.

The issue centered around Apple’s Advanced Data Protection (ADP) tool, which is a feature that uses end-to-end encryption to ensure only users that are signed in on a trusted device can view their personal files, excluding even Apple from accessing the data.

The order issued by the government, which will not be published, would have required Apple to give authorities a backdoor through which it could access the encrypted data for UK users. Apple has consistently reiterated that it opposes creating any backdoors for its encryption because it defeats the entire purpose of encryption – so it opted to simply turn off ADP in the UK, meaning that not all customer data stored via iCloud will be fully encrypted.

I believe the UK government’s decision is a total travesty, both from the perspective of global privacy and cybersecurity, as well as its ability to compete in the global tech industry.

Firstly, from a consumer rights perspective this is an unquestionable loss. Though the government maintains that encryption backdoors are important for combatting crime this logic totally misses the myriad ways in which encryption protects citizens.

Modern cryptography techniques have been a huge leap forward in terms of protecting people’s private data. Encryption means you can access your bank account from your phone securely, or view medical records online without worrying about cyber criminals intercepting your communications.

The bottom line is that in the internet age encryption is vital for almost all digital services and the continued attempts by governments to ensure they have a master key not only violates privacy, but undermines the overall security of the internet.

UK demands could set privacy and security back decades

What the government fails to understand is that compromising the integrity of end-to-end encryption will introduce a raft of new harms for individuals and businesses alike.

As previously stated, introducing a backdoor to break encryption defeats the purpose of the entire exercise and gives malicious actors an easy target for compromising the most sensitive data.

The UK government has made similar attempts at circumventing encryption layers used by private companies in the past as noted in a 2015 paper published in the Journal of Cybersecurity looking at the security implications of requiring ‘exceptional access’ to information systems.

In the paper, researchers perfectly laid out why government efforts to access information would hinge on developers coming up with designs that minimized the impact of breaches, but that this is “unworkable in practice” and would set the global cybersecurity community back decades.

“The complexity of today’s Internet environment, with millions of apps and globally connected services, means that new law enforcement requirements are likely to introduce unanticipated, hard to detect security flaws,” the researchers wrote.

The UK government is setting a dangerous precedent by insisting its agencies be given a backdoor to private companies’ encryption systems. It frequently lambasts other governments such as China for privacy violations, but the latter still allows Apple’s ADP technology in the region.

I’m far from alone in my feelings of disappointment in the government for pushing ahead with its decision. Experts from across the tech sector have argued that undermining encryption is unacceptable.

In an unpublished letter shared on X, Meredith Whittaker, president of the Signal Foundation, which has warned it may have to quit the UK if it continues to undermineend–to-end encryption, wrote that the move puts sensitive information belonging to UK businesses and citizens at “grave risk from malicious actors”.

“Technically illiterate” request risks harming UK’s tech ambitions

The current government follows the last in citing its ambition for the UK to become a global tech hub and unlock AI opportunities. But by taking actions like this, it is seriously shooting itself in the foot.

Legal experts such as Will Richmond-Coggan, a partner specializing in cybersecurity and privacy disputes at Freeths LLP, said Apple’s decision to turn off ADP could have a domino effect on other major tech companies, warning “the UK may no longer be seen as a safe destination for personal data”.

“If that in turn results in the UK losing its adequacy status with the EU, every company doing business in Europe will be subject to additional costly compliance obligations, hampering the government's plans for growth, and a closer European political relationship.”

Whittaker described the move as “technically illiterate” restating that Signal would sooner stop operating in the UK than compromise its encryption.

“The UK's technically illiterate demand also undermines its ambitions to become a tech hub. You can’t be tech friendly while eroding the foundation of cybersecurity on which robust tech depends. Encryption is not a luxury – it is a fundamental human right essential to a free society that also happens to underpin the global economy,” she argued.

This furore all started with the UK government taking exception to an optional feature that the majority of users don’t have turned on – making a mountain of a molehill – but it could have major consequences.

The government should roll this back, and quickly, if it wants to send the right signals to global business and regain any credibility it has on discussions on privacy and cybersecurity.

Apple CEO Tim Cook has hailed the company's "deep connection" with the UK as it continues its expansion efforts across the country.

Cook's comments follow the publication of figures by Apple showing it has doubled the size of its engineering workforce in the UK over the last five years.

The tech giant has also invested more than £18 billion in the country during that time.

"We’ve been serving customers in the UK for more than 40 years, and we’re proud of our deep connection with communities across this country," Cook said.

"We’re thrilled to be growing our Apple teams here, and to keep supporting the extraordinary innovators, creators, and entrepreneurs who are pushing the boundaries of technology in so many ways."

The tech giant now supports 550,000 jobs through direct employment, its supply chain, and the iOS app economy, according to a recent announcement.

Its engineering teams in London, Cambridge, and elsewhere are working on support for key technologies in Apple Intelligence, such as Siri’s deep integration into Apple products, and Private Cloud Compute.

There are also UK teams working on silicon engineering, along with services like Apple Music and Apple TV+, where production has tripled in the last two years alone.

According to Apple, the UK boasts the largest developer community in Europe.

Since the launch of the tech giant’s App Store in 2008, developers across the country have earned nearly £9 billion from selling digital goods and services, it said

The company also highlighted its educational programs, including The King’s Trust, which it has since helped to expand its STEAM programming to every region across the country through the organisation’s Achieve programme.

Apple continues its UK expansion

Last year, Apple launched a major recruitment drive in the UK, looking specifically for AI development staff around the country. It currently has more than 150 positions open at various locations around the country.

It also opened a new office in Cambridge focused on AI and machine learning with several hundred staff, as well as a six-storey, 500,000 square foot headquarters in London at Battersea Power Station.

It's not all plain sailing in the UK for the company, though. It's currently facing a class action lawsuit for charging UK developers a 30% fee.

The case was last week given the green light by the Competition Appeal Tribunal (CAT) and could see Apple forced to pay back fees to 13,000 developers amounting to up to £785 million.

"Apple’s App Store is currently the only means by which app developers can distribute their apps on iOS devices.," said competition law expert Sean Ennis, who is leading the claim.

"Apple exploits this fact by charging excessive commission fees that do not reflect the significant value that third-party apps contribute to the success of iOS devices."

"Apple’s exploitation of its monopoly on iOS app distribution deprives developers of fair compensation and funds that could be better spent on research and development."

Apple has unveiled the new Mac Mini sporting an overhauled and even smaller chassis with souped-up internals, promising the latest version of its compact desktop is “small but fierce”.

The new model features Apple’s much-lauded M4 and M4 Pro chips, which is significant considering the Mac Mini did not get an M3 version, so the performance boost should be dramatic.

Apple said users can expect the new Mac Mini to deliver 1.8x faster CPU performance and 2.2x faster GPU performance over the M1 model.

All versions of the new Mac Mini will now ship with 16GB of RAM, with 8GB now deemed insufficient for modern workloads.

Apple also noted the M4 and M4 Pro chips will support hardware-accelerated ray tracing for the first time, which will be a boon for creatives.

Apart from the silicon, the biggest change is the size of the new Mac Mini, coming in at just five-by-five inches - a footprint that is less than half the size of Apple’s previous design.

Apple said it has implemented a new air flow system and “innovative thermal architecture” in the design to accommodate thermal throttling concerns with the limited space.

Apple’s new Mac Mini goes carbon neutral

The new Mac Mini marks an important milestone for Apple, being the first carbon neutral Mac with an over 80% reduction in greenhouse gas emissions associated with materials, manufacturing, transportation, and customer use.

The latest Mac Minis will offer a slightly altered port lineup too. The front has a 3.5mm headphone jack and two USB-C inputs, with a gigabit ethernet, HDMI, and three Thunderbolt 4 ports round the back.

Users that spring for the more expensive M4 Pro model, will enjoy higher data transfer speeds with the Thunderbolt ports on the rear upgraded to Thunderbolt 5.

The M4 Mac Mini will start at £599 for 16GB of RAM and a 256GB SSD and £799 to double the storage capacity to 512GB.

The M4 Pro version, with a 12-core CPU and 16-core GPU, will start at £1,399 and will come with 24GB of RAM and a 512GB SSD.

Apple is capitalizing on the burgeoning mini PC market

It’s been well over a decade since Apple last gave the Mac Mini a design overhaul, and it appears to be doubling down on the ‘mini’ part of its name. The timing of the announcement and the size of the machine could be seen as strategically shrewd by Apple.

The refresh also includes an M4 chip, bringing the Mini in line with both Apple’s MacBooks and iPad ranges. And, also positioning its product near the top of the mini PC market, which is seeing significant growth through gaming, small business adoption, digital signage, and remote work.

The market for mini PCs was valued at $21.1 billion in 2022 and market reports suggest its continued growth could reach $33 billion by 2032.

Apple is offering a $1 million (£770,000) bounty for flaws that could be used by hackers to run code in Private Cloud Compute (PCC), the cloud system used to power advanced AI features that can't run on device.

Bug bounty programs are a popular way for technology companies to incentivise security researchers to seek out vulnerabilities in their code, offering rewards for any flaws found. Last year, Google paid out more than $10 million on bug bounties to more than 600 researchers, including its largest ever reward at the time worth $605,000.

Apple unveiled Apple Intelligence, its system for generative AI on iPhones, iPads and Mac computers, earlier this year. While some AI features process data on devices for privacy reasons, advanced features that require more processing power are handled by Apple's new cloud, Private Cloud Compute.

Apple says it designed its cloud system to have the same levels of security and privacy as Apple devices, including not allowing Apple itself to view the data, in order to help build user trust in AI — or at least AI on Apple devices.

To help Apple spot vulnerabilities in those cloud servers, it has extended its pre-existing bounty program to cover PCC — with big rewards for the most dangerous flaws.

"To further encourage your research in Private Cloud Compute, we’re expanding Apple Security Bounty to include rewards for vulnerabilities that demonstrate a compromise of the fundamental security and privacy guarantees of PCC," the company said in a post outlining the bug bounty.

Apple noted that the bounties are in line with those for iOS — which makes sense as PCC is pushed as an extension of a user's device.

For iOS, Apple also has bounties stretching from $5,000 to $1 million, and even offers bonus payouts that could take rewards to the $2 million mark for iOS flaws in new features in beta software or that target lockdown mode.

Apple's bountiful bounties

The new PCC bounty rules are centered on what Apple sees as the most serious threats: accidental data disclosure, external compromise from user requests (such as flaws that allow hackers to exploit the system), and physical or internal access.

"We award maximum amounts for vulnerabilities that compromise user data and inference request data outside the PCC trust boundary," the post added.

Apple splits the PCC bounties into two main categories. It will pay rewards from $50,000 to $150,000 for attacks from a privileged network position, with the highest payouts for vulnerabilities that can access a user's request data or other sensitive information.

The biggest bounty is for the second category, remote attacks on request data. Flaws that would allow access to data could earn a bounty up to $250,000 while those that allow code execution could be rewarded with up to $1 million.

Apple added that it still wants to hear about flaws that land outside these categories, of course.

"Because we care deeply about any compromise to user privacy or security, we will consider any security issue that has a significant impact to PCC for an Apple Security Bounty reward, even if it doesn’t match a published category," the post added.

As most IT departments know, Apple MacBooks are highly sought-after laptops. Even in this age of NPU-powered Windows machines, the mighty Mac still holds great appeal in the office.

And, the MacBook Air, with its slimmer chassis, is a lightweight option for those who roam. It comes with fewer port options but it's great for those who work remotely or move around different workspaces in the office. Our review unit has a 13.6in display, the Apple M3 chip with the 8-core CPU and 10-core GPU, 16GB RAM, and 512GB storage – those can all be configured to suit, for a higher fee.

Apple M3 MacBook Air 13.6in: Design

MacBooks seldom change in appearance and there isn't much to get excited about here, aside from a new shade of navy, called 'Midnight', which is rather lovely. But visually there isn't much difference between the M3 and M2 models of MacBook Air.

The overall build is as good as ever with a recycled aluminium chassis and a pleasingly stiff lid. It's the first Apple product made up of 50% recycled materials, which is good. However, you can't upgrade it – there's no easy access to replace any parts, so it's not all that eco-friendly. It is, however, a very slim device with just 11.3mm thickness when closed and it's light at 1.24kg – perfect for a rucksack.

Almost all of the new design elements are on the inside, such as the M3 chip and support for Wi-Fi 6E. The keyboard, trackpad, and display are all largely the same as the M2 MacBook Air. It's the same iconic design that we know and love.

Apple M3 MacBook Air 13.6in: Display

There's no real change to the display, either, as the M3 MacBook Air has Apple's 'Liquid Retina' panel, which is a fancy name for a quality IPS screen. It measures 13.6in diagonally, supports P3 color reproduction, and comes with a sharp 2,650 x 1,664 resolution. Sadly there is only up to 60Hz for the refresh rate which is dwarfed by most of its Windows-based rivals that are hitting 120Hz. It is also worth noting machines that use more vibrant OLED panels where you can get more from the display, for a lot less.

In our colorimeter tests, the MacBook Air showcased 97.1% of the DCI-P3 gamut, which is a high score and a testament to its stunning visuals – though other screens, such as the Lenovo Yoga Pro 9i 16, make this look dull (101.9%). There is also less brightness here than the M3 MacBook Pro, though not enough to be concerned.

What is concerning, however, is the default color settings. Out of the box, the MacBook Air has "Color LCD" mode, which is the main one used by Apple devices. Here, the gamut is restricted to sRGB colors, making it duller than all the other setting options. But, if you switch to the actual sRGB profile, it does the opposite and offers a more vivid spread of colors. This is probably just a small oops that will be fixed in a future update and nothing to really put you off this model, but it's strange, all the same.

Apple M3 MacBook Air 13.6in: Keyboard and trackpad

You get a great keyboard experience on the MacBook Air, with nicely cushioned keys and decent travel. It suits those who type at high speed with its gently muffled clack and comfortable layout. The keys are all quite large and well-lit, so they're easy to hit. We're also big fans of the large power on/fingerprint reader in the top corner of the keyboard which works seamlessly.

The Force Touch touchpad is as good as ever – arguably the best around. Largely because it is so big and expansive, but also with its haptic feedback. You don't get a physical click here, but it somehow makes you feel like there is one. There's even the slight clack sound – it's just a nice user experience.

Apple M3 MacBook Air 13.6in: Specs and performance

In case you missed it, the main addition to the 2024 MacBook Air is the M3 chip. This is a 8-core processor, built with Apple's innovative 3nm manufacturing process. While that sounds impressive, the result is only a small bump up in performance from the M2 chip of the previous Air models.

In Geekbench 6, the MacBook Air hit 3,085 for single-core, and 11,715 for multi-core – again, impressive. Though it is slightly behind its rivals on multi-core performance. The Acer Swift 14 AI, for example, hit 13,029 with its 10-core Snapdragon X Elite chip. What's more, if you're looking for MacBooks for your workforce, it's worth taking a look at the M2 Air as it offers near similar scores and you can get it cheaper.

You also get the slightly better battery; in our looped video test the M3 Air lasted 14hrs, and 33mins, almost three hours behind the M2 MacBook Air. Admittedly, that is a minor decrease that doesn't necessarily translate to real-world use. Our review unit was put through its paces for a whole week and showed absolutely no signs of stress, offering barely a hum of strain while processing all we could throw at it – but that's an extra $100 for basically the same performance.

Apple M3 MacBook Air 13.6in: Features

One big difference with the M3 MacBook Air is a feature that lets you power two external monitors. You have to close the lid for this to work properly, which is annoying. However, the switch between the Air's display and the second monitor once that lid closes is rapid – you will struggle to find a Windows machine that can keep up.

The port selection on the MacBook Air could be better tough. You have a 3.4mm headphone jack and two Thunderbolt 4 slots on the left-hand side, next to the MagSafe power connector. That's it. If you want anything else you have to stump up the cash for the MacBook Pro and this is probably how they keep the Air so thin and light, but it is still a slight bugbear for us.

Apple M3 MacBook Air 13.6in: Is it worth it?

So, in summary, everything that sets the M3 MacBook Air apart from the M2 model equates to about $100 dollars more. That's a slightly faster processor, the ability to connect two external monitors and the dark blue color option – as good as it is, it makes the M2 MacBook Air look more appealing.

However, if your business has reached its laptop upgrade cycle – or you're a new business ready to kit out employees with an attractive machine – the M3 MacBook Air is light, snappy, and stylish.

Apple M3 MacBook Air 13.6in specifications

Researchers at Cisco Talos have uncovered serious vulnerabilities in Microsoft applications for the macOS operating system that could allow attackers to misuse permissions. 

The vulnerabilities can be exploited by injecting malicious libraries into Microsoft applications to gain entitlements and user-granted permissions.

These permissions control whether an app can access resources such as the microphone, camera, folders, screen recording, user input and more, allowing attackers to spy on the user or steal sensitive information.

"We identified eight vulnerabilities in various Microsoft applications for macOS, through which an attacker could bypass the operating system’s permission model by using existing app permissions without prompting the user for any additional verification,” researchers said.

“If successful, the adversary could gain any privileges already granted to the affected Microsoft applications. Microsoft considers these issues low risk, and some of their applications need to allow loading of unsigned libraries to support plugins and have declined to fix the issues."

Apple's macOS features a layered security model, including Transparency Consent and Control (TCC) – a framework developed by Apple to manage access to personal data and system privileges, requiring explicit user approval before granting access.

However, this isn’t foolproof, researchers warned. It depends on applications responsibly handling the permissions they receive. If a trusted application is compromised, it could be manipulated to abuse its permissions, allowing attackers to perform actions without user knowledge.

"For instance, if a video chat app with camera and microphone access is exploited, it could be forced to record without alerting the user," Cisco Talos said.

"This situation points to a key aspect: macOS trusts applications to self-police their permissions. A failure in this responsibility leads to a breach of the entire permission model, with applications inadvertently acting as proxies for unauthorized actions, circumventing TCC and compromising the system's security model."

As an example, Cisco Talos used Microsoft apps, each of which had hardened runtime enabled, together with the ‘com.apple.security.cs.disable-library-validation’ entitlement.

Of the eight applications it reported, four were updated by Microsoft and no longer possess the 'com.apple.security.cs.disable-library-validation' entitlement.

These included: Microsoft Teams, WebView.app (Microsoft Teams helper), com.microsoft.teams2.modulehost.app helper, now renamed Microsoft Teams ModuleHost.app, and Microsoft OneNote.

However, Microsoft Excel, Outlook, PowerPoint, and Word all remain vulnerable.

Microsoft appears to use the ‘com.apple.security.cs.disable-library-validation’ entitlement for certain apps to support certain plug-ins. According to Apple, this allows the loading of plug-ins signed by third-party developers.

Despite this, the only plug-ins available to Microsoft's macOS apps appear to be web-based and known as Office add-ins.

"If this understanding is correct, it raises questions about the necessity of disabling library validation, especially if no additional libraries are expected to be loaded. By using this entitlement, Microsoft is circumventing the safeguards offered by the hardened runtime, potentially exposing its users to unnecessary risks," Cisco Talos said.

"It's also important to mention that it’s unclear how to securely handle such plug-ins within macOS' current framework. Notarization of third-party plug-ins is an option, albeit a complex one, and it would require Microsoft or Apple to sign third-party modules after verifying their security."

The company recommends Apple should introduce a user prompt, similar to the resource permissions in TCC, enabling users to decide whether to load a specific third-party plug-in.

This, it said, would provide a more controlled means of granting access without broadly compromising security.

Apple has laid out how it trained and developed the AI models powering the Apple Intelligence tools set to arrive with iOS 18, stressing the work was done responsibly amid complaints about data snatching to train generative AI models. 

Leading AI developers have been accused of dodgy tactics when it comes to harvesting data for developing and training large language models (LLMs), prompting a series of lawsuits from newspapers and authors targeting OpenAI and other companies for copyright infringement.

Last month, Apple announced it would be using its own AI models alongside OpenAI's technology to power generative AI tools in the next version of its mobile operating system, iOS 18, due out this autumn.

The first look at that system was unveiled at the 2024 Worldwide Developers Conference (WWDC) in a developer beta for iOS 18.1, suggesting the AI tools may not be immediately available on iOS 18.

Apple’s responsible AI work

At WWDC, Apple also released a report detailing how two of its foundation language models for Apple Intelligence work — and beyond that, laying out the case for why its AI is responsible with regard to data use. The two models in question — Apple Foundation Model (AFM)-on-device, and AFM-server — are part of a larger suite of generative models created by Apple. 

"Apple Intelligence is designed with Apple’s core values at every step and built on a foundation of industry-lead privacy protection," the document states. "Additionally, we have created Responsible AI principles to guide how we develop AI tools, as well as the models that underpin them."

Those principles guide Apple to identify where AI can be used responsibly to help users; avoid bias; take care to avoid harm at all stages of design and development; and protect user privacy.

"Apple Intelligence is developed responsibly and designed with care to empower our users, represent them authentically, and protect their privacy," Apple said via the report.

In practice, that means Apple methodically tests its AI against a list of possible concerns, such as hate speech and graphic violence when it comes to model training, guardrail development, evaluation, and data collection, the company said.

Where's the data from?

Apple said AFM pre-training data came from "a diverse and high quality data mixture" — and includes no private Apple user data. 

"This includes data we have licensed from publishers, curated publicly available or open-sourced datasets, and publicly available information crawled by our web-crawler, Applebot," the report said. "We respect the right of webpages to opt out of being crawled by Applebot, using standard robots.txt directives."

That comes as AI developers including Anthropic have been accused of scraping websites for content to train with despite instructions in the page — the robots.txt file that Apple references — that ask them not to. While the practice isn't necessarily illegal, it can breach copyright and cause disruption and high costs to websites.

Meanwhile, Apple's reference to publicly available datasets could include "The Pile". Apple is one of several companies accused of using YouTube video subtitles in its training after reportedly using a publicly available research dataset known as the "the Pile" to train its OpenELM model.

That suggests that even with such responsible data gathering principles in place, data from less than suitable sources could still be used by Apple.

Beyond text, Apple pulls in code sourced from GitHub, stressing it uses only code data with licenses that allow such reuse — clearly attempting to avoid a lawsuit like that faced by Microsoft and GitHub over the use of open-source software.

Synthetic data is also used in the post-training phase, in particular with coding, math, and tools.

Apple added it cleans up all data — "extensive efforts have been made to exclude profanity, unsafe material, and personally identifiable information" — regardless of the source, be it public datasets or licensed from publishers.

Is Apple the ethical AI alternative?

It's clear that Apple is attempting to set out its stall as the responsible AI provider, just as it has done with privacy on mobile. 

That could be a wise move. Research suggests a distrust of generative AI and the companies that make the systems. Plus, the ongoing legal wrangling around data could prove a challenge to rival systems, especially if models need to be reworked in the future to avoid using specific datasets.

The ethical effort could pay off with better results in the long run, if the tech axiom "garbage in, garbage out" proves true with LLMs.

"We find that data quality, much more so than quantity, is the key determining factor of downstream model performance," Apple said in the report.

Microsoft employees in China have been told they must ditch Android smartphones in favor of Apple iPhones over security concerns. 

To be clear, Microsoft isn't suggesting that smartphones running Google's Android mobile operating system are inherently insecure. Instead, the swap to iPhones stems from China banning the Google Play store, limiting Microsoft employees' ability to download key verification apps.

Apple's App Store, however, is still available.

According to a memo seen by Bloomberg, Microsoft wants its staff to make use of its Identity Pass app and Authenticator password manager when accessing corporate networks, requiring access to an app marketplace.

The report said the move will impact hundreds of Microsoft employees in mainland China, though Microsoft will be supplying the new handsets.

Microsoft confirmed the move and motivation in a statement sent to journalists.

"Microsoft Authenticator and Identity Pass apps are officially available on the Apple and Google Play stores,” the firm said. “Due to the lack of availability of Google Mobile Services in this region, we look to offer employees a means of accessing these required apps, such as an iOS device."

As other publications have speculated, it's certainly possible for Microsoft to side-load these apps or use local non-Google app markets, such as Tencent. But instead, Microsoft has chosen to simply buy iPhones.

Perhaps that's little surprise given Microsoft has closed its shops in the country, and is reportedly encouraging cloud and AI staff to relocate, amid worsening relations between the US and China.

Ilia Kolochenko, CEO at ImmuniWeb, said the move could also be due to lingering concerns about the safety of open source software products.

"I think this move may be at least partially triggered by the growing lack of Western confidence in open-sourced products, bolstered by the recent XZ Utils backdoor scandal," Kolochenko said.

Microsoft’s latest move to bolster security

The move fits in with Microsoft's wider efforts at securing its own operations under a plan known as the Secure Future Initiative (SFI). 

First unveiled in November last year, with an update seen this May, SFI is in response to the "increasing scale and high stakes of cyber attacks," according to executive vice president Charlie Bell in a post in May.

SFI is essentially an effort by Microsoft to shore up security for itself and its customers under a secure by design, by default and by operations framework.

In May, Bell noted that was being expanded to include new priorities, among them protecting people and data using authentication and authorization tools, setting a target for all user accounts and applications to be protected via multi-factor authentication.

While Microsoft's shift from Android to iPhone in China is specifically to do with access to these verification apps, Android has long been a major target for hackers. 

That's in part down to the open Android ecosystem — security risks can be introduced by the myriad hardware manufacturers, or by installing apps from third-party marketplaces, for example.

On the other hand, iOS has long been seen as more secure, partially down to Apple's more locked-down approach, not only manufacturing the phones itself but also constraining access to core code — though it's worth noting that no smartphone offers perfect security against hackers.

“In America, Apple has a better perception of security and privacy compared to Android, although it may not necessarily be true: a properly configured and hardened Android is much more secure than an outdated iOS device,” Kolochenko noted.

Regardless of the phone platform, apps are a key target for hackers, which may be one reason why Microsoft would prefer not to sideload its authentication apps or ask workers to use alternative marketplaces.

Virtually every single Apple device on earth was exposed to a number of critical vulnerabilities through the dependency manager CocoaPods, new research has revealed.

CocoaPods is a popular open-source dependency manager for Swift and Objective-C used by Apple developers to manage external libraries, with around 100,000 libraries used in over three million mobile apps.

EVA Information Security revealed it found several vulnerabilities in CocoaPods that could allow an attacker to claim ownership of potentially thousands of unclaimed pods and execute arbitrary code on the service’s trunk server.

According to the investigation, the vulnerabilities arose in 2014, and were only patched in October 2023, lying dormant for nine years waiting to be exploited.

The first of these vulnerabilities arose after a 2014 migration to a new trunk server left thousands of orphaned packages, meaning the original owner is unknown, although many of them are still being used in other libraries.

This meant attackers could use a public API to claim the pods without any verification required, with almost 2,000 pods still unclaimed by their owners, and waiting to be claimed by threat actors.

These pods could be injected with malicious code and used in supply chain attacks that could compromise potentially millions of iOS and MacOS devices around the world, EVA team warned. 

This flaw, CVE-2024-38368, designated a 9.3 on the CVSS, also allowed attackers to remove all owners from any pod, making them available to be claimed by malicious parties too.

The second vulnerability EVA flagged, CVE-2024-38367, is a session hijacking flaw that leverages a loophole in CocoaPods’ authentication process allowing attackers to manipulate the verification process leading to a full takeover of the CocoaPods trunk account.

EVA’s investigation stated that CocoaPods authenticates new devices via email, requiring the user to simply verify their email address by clicking a link, but this functionality could easily be hijacked, the report warned.

“We found that the server will accept a spoofed XFH header and use it explicitly to construct a URL sent to the client for verifying the session.” EVA researchers said.

EVA noted that with a few extra steps attackers could turn this into a zero-click account takeover attack. Attackers can rely on the fact that most email servers feature  automated phishing detection tools that scan potentially dangerous links.

But when scanning the spoofed CocoaPods validation link, the tools inadvertently transmit the session token to the attacker, which they can use to validate their session and take control over the account and its packages.

Critical flaw could've spelled disaster for Apple devices

Changes to the CocoaPods trunk source code to create a new email verification workflow also created a new attack path, EVA reported, opening the door for the third, and most serious, vulnerability.

Rated 10 on the CVSS, CVE-2024-38366 is a remote code execution (RCE) flaw that would enable an attacker to execute arbitrary code on the trunk server.

A change committed to the trunk source code implementing MX record validation to registered emails, utilizing a third-party ruby gem package rfc-822, which unfortunately has a number of vulnerable validation methods.

As a result, EVA explained an attacker could exploit any of these validation methods to execute commands on the trunk server.

Although the report noted there is no ‘direct’ evidence of any of these bugs being actively exploited in the wild, it reminded readers “evidence of absence is not absence of evidence”, listing a number of mitigation steps organizations can follow to ensure their apps are not compromised.

Developers who have released applications for Apple devices should keep their podfile.lock file synchronized with all CocoaPods developers, ensuring that if a harmful update is committed, developers will not automatically update to it.

Developers using a pod developed internally and only hosted in CocoaPods for mass distribution should perform CRC validation against the one downloaded from the CocoaPods trunk server to guarantee it is the same as the one developed internally.

Finally, EVA recommends caution of very widely used dependencies as they could be the first targets the attackers try to exploit to achieve maximum yield with their attack.

Swift is the best programming language to replace C++, according to a senior figure at Apple, as scrutiny over software safety continues to rise. 

Ted Kremenek, Apple director of languages and runtimes, made the claim at Apple’s World Wide Developer Conference (WWDC) earlier this month while announcing a raft of updates for the programming language.

Introduced by Apple in 2014, Swift was originally intended to act as a successor to Objective-C. However, the popularity of the programming language has boomed in recent years, and over one million applications are now using Swift.

Kremenek noted that the language has now become crucial for Apple, and the company has been keen to promote it as a safer alternative to C++.

“It’s fast, modern, and safe, and a joy to write,” he said.

“We created swift to be an expressive and safer programming language that would simplify the process of writing software,” Kremenek added. Swift is an ideal language for app development, and is used by almost one million apps.”

Swift is great for “more than just apps”, however. Kremenek said Apple now uses Swift throughout its software stack, “from apps and system services to frameworks all the way down to firmware, like Secure Enclave”.

The programming language is also used in network services like Apple’s Private Cloud Compute (PCC).

“Combined with built-in C and C++ interoperability means Swift is the best choice to succeed C++.”

Why Apple is championing the shift to Swift

Prior to the advent of Swift, software on Apple devices was primarily written using C, C++, and Objective-C, Kremenek noted. In recent years, however, there’s been a degree of pushback against languages such as C and C++, specifically due to security concerns. 

In November 2022, the US National Security Agency (NSA) recommended only using ‘memory safe’ languages like Rust, Swift, or Java to avoid exploitable memory-based vulnerabilities.

This initial call to arms has since gathered pace amid rising concerns over software vulnerabilities across the last 18 months. In February this year, the White House once again urged developers to ramp up efforts to switch to memory safe languages.

The warning followed the publication of a report by the White House Office of the National Cyber Director (ONCD), which found that developers can prevent entire classes of software vulnerabilities by adopting safer options.

The White House has a vested interest in banging the drum for memory safe languages given the critical importance of software and hardware manufacturers in bolstering national cyber resilience.

Like lawmakers, Apple’s contention here is that moving to Swift will improve the quality of software not only at the company but throughout the industry.

Apple is committed to Swift

Apple has made no secret of its transition to Swift in recent years, and during his keynote at WWDC Kremenek teased plans for the programming language. 

Chief among these are plans to launch a new GitHub organization dedicated to Swift, github.com/swiftlang. This will host “a number of critical projects” for the Swift ecosystem, including the Swift compiler, foundation, and other “key libraries”.

Kremenek also made reference to the launch of Swift 6, which is expected to arrive later this year. Swift 6 will make concurrent programming “dramatically easier” by introducing “data race safety”, he said.

A data race occurs when different parts of code try to modify and access the same data simultaneously, Kremenek explained. With Swift 6, however, these kinds of bugs will be eliminated by diagnosing them at compile time.

For some time now, Apple has been viewed as something of a sleeping giant in the generative AI space. While Google, Microsoft, and a myriad of big tech companies have been slugging things out in this domain, the company has been holding its cards close to its chest. 

It’s a talking point that hasn’t gone unnoticed over the last year, and the firm has been quietly drip feeding hints at its plans. In December, Apple launched a range of open source tools in a bid to provide on-device machine learning capabilities.

The Apple MLX framework allows users to build AI tools harnessing Apple Silicon, and was touted as a major hint at its pending AI ambitions. This was followed by speculation in February that the tech giant was building an AI coding tool similar to GitHub Copilot.

CEO Tim Cook has also offered up frequent veiled hints on the company’s AI ambitions over the last year, but nothing truly groundbreaking has been announced so far.

That’s all about to change, however. At its Worldwide Developers Conference (WWDC), Apple firmly planted its flag in the generative AI space, unveiling a raft of new AI-powered tools and features, as well as a major partnership with OpenAI to embed ChatGPT within Siri.

Apple’s marquee announcement centered around the launch of Apple Intelligence, the company’s “personal intelligence system” that integrates generative AI capabilities across iPhone, iPad, and Mac devices.

In quintessential Apple style, Cook was eager to emphasize that safety and data privacy will be at the core of Apple Intelligence features.

“It has to understand you and be grounded in your personal context, like your routine, your relationships, your communications, and more,” he said.

“And of course, it has to be built with privacy from the ground up. All of this goes beyond artificial intelligence,” Cook added. “It’s personal intelligence, and it’s the next big step for Apple.”

Apple Intelligence will be woven into the very fabric of the company’s flagship operating systems, the company revealed, meaning iOS, macOS, and visionOS users will have access to new features.

All of these new capabilities are very much focused on making life easier for users. For example, people with iPhones will now be able to record, transcribe, and summarize audio using Apple Intelligence.

Image tools have also been added across the company’s platforms, allowing users to create emojis or illustrations by describing them. Even the calculator app has a dynamic new look and will be able to solve equations for users in real-time.

However, these new features do require significant processing power. But the company has compensated for this. Apple confirmed it is adding Private Cloud Compute to its offering. This essentially allows users to offload complex AI tasks to the cloud, rather than processing them on-device.

This has two key advantages, the company said. Users will be able to harness private AI processing capabilities while bolstering privacy. Apple described PCC as the “most advanced security architecture ever deployed for cloud AI compute at scale”.

“With groundbreaking Private Cloud Compute, Apple Intelligence can draw on larger server-based models, running on Apple silicon, to handle more complex requests for you while protecting your privacy,” the firm said.

Apple’s shiny generative AI features are more than just a consumer market grab

All the announcements at WWDC will likely excite Apple customers, but there’s more to this latest move than just flashy new tools. This marks the company’s first major foray into the generative AI space - and one that firms such as Microsoft will likely have been keeping a close eye on throughout the keynotes. 

Thomas Husson, VP principal analyst at Forrester, said that while many of the announcements were expected, what really matters here is how Apple will utilize its developer ecosystem to drive continual improvements and deliver tangible benefits for users.

“Taken in isolation, each announcement looks like a minor incremental innovation to improve the user experience,” he said. “What matters is the total sum of improved experiences that brands will be able to develop on top of devices across Apple’s ecosystem, and how consumers perceive them.

“Apple has always been paying attention to detail and telling a story that benefits people in their daily lives. Today, Apple resisted constantly referring to the overhyped GenAI buzz and highlighted that they have been investing in many different AI technologies, such as machine learning, for quite some time.”

Anisha Bhatia, senior technology analyst at GlobalData, said the announcements showed that Apple is “no longer playing catch-up” in the generative AI race, with its news features pitting it firmly against Google and Microsoft.

“The company is off to the races with its cheeky branding of ‘Apple Intelligence’ and a horde of on-device features, on par with Google and Microsoft in the generative AI arena,” Bhatia said.

“Apple will be rolling out its intelligence features systemwide and across its own apps in iOS 18 for iPhones, iPadOS 18 for iPads and Mac Sequoia for Macbooks, showcasing the powerful role played by Apple's ecosystem in its AI implementation.”

Both Bhatia and Husson highlighted the fact that Apple deliberately avoided cumbersome technical jargon throughout the keynotes, which will likely have been a breath of fresh air for consumers incessantly bombarded with flamboyant claims over the last 18 months.

Messaging, in Apple’s case, will be key here given its late arrival to the arena. But in calmly articulating its vision and outlining the tangible, every day benefits of AI, it may have already struck an early lead in the marketing battle that will unfold in the coming months.

“Unlike its competitors, Apple has taken a consumer-friendly approach to explaining the benefits of AI,” Bhatia said.

“The emphasis on on-device intelligence and Private Cloud Compute aligns with Apple's narrative around privacy and transparency. Android rival Google’s marketing and messaging around AI has been scrambled at best.”

The Apple M3 MacBook Pro came as a surprise at the end of 2023, particularly to those who had picked up an M2 MacBook much earlier in the year. The pace of change is relentless as we are already talking about M4 chips in iPads, but it is worth holding on to M3 for a moment, if only just to appreciate the sheer quality of this latest MacBook Pro. 

It comes in a wide range of price points and specifications. As well as a slick new shade – 'Space Black'. The basic option is $1,849 for a 14in model but you can max it out to 16in and 8TB storage, though that will cost you over $7k. For this review, we've got a more modest 14in Pro M3, with 18GB of RAM and 1TB of storage, but it has at least come in that fancy new color. 

Apple M3 MacBook Pro review: Design

At face value, there isn't a lot of change for the new MacBook, aside from the lovely black finish. It is almost exactly like the M2 model, which isn't a problem, as that still means you get an attractive laptop with a solid aluminum chassis. The lid is particularly pleasing as it is quite stiff and sturdy. 

The design is solid, overall, and there is a great selection of ports. You get a full-size SDXC card slot, HDMI 2.1 output, and a MagSafe charging port. There are two USB-C ports on the left-hand side of the machine, which support Thunderbolt 3 and USB 4 data transfers, charging, and DisplayPort video output. Understandably, those ports bulk out the MacBook Pro, resulting in a chunky chassis that weighs 1.55kg. But this isn't much bigger than you would expect from a MacBook Pro. 

Apple M3 MacBook Pro review: Display, keyboard, and trackpad

Apple's 14in liquid retina XDR display is very nice. It has a 3024 x 1964 resolution, 254 pixels per inch, and an exceedingly bright panel. We can't say it's the best of the best – it's no Samsung Galaxy Book 3 Ultra – but it is one of the most vibrant and color-accurate screens we've seen on a Mac. One thing that does stand out as a fault is the notch at the top; this was introduced with the M2 MacBook Pro and it really does need to go.   

Otherwise, this is an amazing display. With our calibrator, the M3 MacBook Pro scored 98% for color accuracy and 120% for volume. Both are impressive scores but again, slightly below the impressive Galaxy Book3 Ultra – though it does have a higher peak brightness, 633cd/m2.

The webcam, which is housed in the notch, produces nice, crisp images, and seems adaptable to most lighting conditions. And the six-driver speaker system is very clear. It makes the Pro is great for video meetings, though there isn't anything that's better or worse than previous MacBooks in that regard. 

The keyboard is decent, though it does tend to rattle when you pick up speed. That being said, the keys themselves are nice, large, and well-spaced, so typos are very few and far between. 

The glass-topped Force Touch touchpad is very good; a highly responsive control with plenty of room to swipe. You can click pretty much anywhere on the pad and get the same response – something that isn't always the case with Windows machines. 

Apple M3 MacBook Pro review: Specs and performance

The MacBook Pro review unit we were sent has the Apple M3 16-core CPU and it is as rapid as we have come to expect from the M-series. We also have 18GB of RAM, which seems like a strange amount, and 1TB SSD, which is very welcome.

M3 and the M-series, in general, have a huge advantage that comes from the way they are developed with a standardized architecture that has more and more cores wrapped around. This gives it stellar performance levels, no matter what size and spec set you opt for.

Under GeekBench 5, the M3 MacBook Pro hit 3,184 for single-core and 15,704 for multi-core scores. Here Apple is looming over Windows-based machines as there is nothing that can match the power of M3, except other Apple laptops. Even the brilliant Dell XPS range pales in comparison. 

We gave the M3 MacBook Pro as stringent a test as we could – 50 tabs, intensive video loops, lots and lots of Adobe edits – and it came through them with barely a whimper of complaint. What's more, Apple promises 22hrs of battery life and to be fair it isn't far off if you do the bare minimum with the machine. Under our looped video test, the Pro lasted 15hrs 44mins, which is very impressive – far longer than the 9hrs and 53mins of the Galaxy Book 3 Ultra. Again, no Windows machine can get anywhere near the latest Macs when it comes to battery life.  

Apple M3 MacBook Pro review: Is it worth it? 

There are many spec sets and sizes of MacBook Pro on offer and these include the suped-up M3 Max versions. The higher you go, the more dizzying the prices get. Small businesses can use Apple Financial Services to find a flexible payment plan, which is probably best for newer businesses. You can also work out upgrades and trade-ins here. Similarly, the Apple for Work service can also help with after-sales support. 

However, even with the most affordable variants, the M3 MacBook Pro is a brilliant laptop. It has a premium feel regardless of RAM size or color. If you're after a professional-looking machine, that's innovative and reliable, the Apple M3 MacBook Pro should be top of your list.

Apple M3 MacBook Pro Specifications

Apple has been going about its business in the generative AI space in quintessential style over the past 15 months. While the global technology industry has been utterly saturated with AI-related hype and hyperbole, the company has been working away quietly in the shadows. 

The firm’s quiet approach to generative AI has created a sense of confusion among industry stakeholders, with some previously begging the question of whether it’s focusing on the technology at all.

Tim Cook, CEO at Apple, has intermittently teased details of Apple’s AI focus over the last year. In 2023 he told CNBC that Apple views AI and machine learning as “core technologies” that will be “embedded in every product that we build”. But the initial wave of excitement over Apple’s potential in the space quickly dissipated amid silence from the firm. 

Unlike its big tech rivals, Apple has seemingly been slow to invest in generative AI; but recent developments show this is less a case of sluggish investment and more reflective of its traditional approach to innovation. 

In a February 2024 earnings call, the Apple chief executive said the company will “break new ground” on generative AI capabilities later this year. Exactly when that new ground will be broken remains to be seen, but there are hints at a big push toward AI by Apple from mid-2024. 

While this puts it more than a year behind the competition on paper, Apple’s long been comfortable chasing quality at the expense of being first to market. It’s rare to see Apple adopt a technology first and rarer still that it fumbles product launches when they finally come around. 

Why risk a rushed, calamitous launch fraught with controversy akin to Google's Bard woes when you can methodically pull together a portfolio of products and services to be sprung on the industry and blow away competition?

Cook specifically highlighted this approach during the company’s 2024 Q1 earnings call in February. 

The chief executive told investors the company was working extensively on generative AI internally, but that it has made a conscious decision to pursue a more considered and reserved approach to the technology. It hopes to use this time to produce fine-tuned products ready for consumer adoption. 

Apple's generative AI approach is slow and diversified

If Apple can deliver ready-to-go generative AI products and services, rather than saddle its customers with early-release tools subject to error, it could avoid mistakes such as ChatGPT’s privacy leaks. Recent indications that the firm is investing in a wide range of AI frameworks and tools put it in a good position to hit the ground running on launch.

In December 2023, Apple launched a raft of open source AI and machine learning tools in a bid to offer users on-device AI development capabilities. Dubbed ‘Apple MLX’, this machine learning framework was developed by the firm’s research division. Ordinarily, the launch of this framework would’ve raised eyebrows, but it was announced against a backdrop of Google Gemini-related hype. 

Whether the timing was intentional is unclear, but regardless the news went overlooked by many. In a year’s time, this quiet service launch may be looked back on as the first clear indication of Apple’s AI engine ignition. 

It’s not the only teaser Apple has thrown our way of late. In mid-March, Apple acquired Canadian AI startup, DarwinAI. The company specializes in vision-based technology to automate observation of components manufacturing, Bloomberg reported at the time. 

The acquisition represented the latest startup to join a growing list of purchases by the firm, but so what? Bloomberg noted that, in addition to AI-supported manufacturing capabilities, DarwinAI has also been working on techniques to make AI models faster and smaller. 

Given the launch of Apple MLX in December, which specifically centers around harnessing Apple Silicon for on-device generative AI development, the acquisition could highlight another building block in Apple’s foundational development. 

Apple MM1 is its biggest AI project yet

Within days of the DarwinAI speculation, the industry gained its biggest glimpse into Apple’s generative AI work to date – and it’s this that should have alarm bells ringing for major players such as Microsoft or Google. 

A research paper by Apple engineers, quietly posted online on 15 March indicates the tech giant is working on the development of its own generative AI model, MM1. 

MM1, according to the paper, is a multimodal large language model, which means it is trained on - and capable of working with - both text and images. 

The research paper details instances in which the model answered questions about specific images and provided answers to queries in a fashion similar to ChatGPT. Based on early signs, it appears capable of going toe-to-toe with other models such as Meta’s Llama 2 or Google Gemini. 

The launch of an in-house multimodal LLM by Apple would represent a step change in the company’s current AI trajectory and position the company in direct competition with Google or Microsoft. 

But again, industry and consumers alike will be left waiting with bated breath for any further details on this, with Apple remaining firmly tight-lipped about its activities in this domain. 

There’s nothing wrong with the strong, silent type

There’s no doubt that this approach is frustrating. Businesses that have invested in the Apple ecosystem want to know what the company is working on in terms of potential AI features in upcoming devices, particularly as rival platforms settle in with AI tools of their own.

But this is a breath of fresh air for the sector. Since November 2022 and the release of ChatGPT, the industry has been bombarded with news on generative AI and there’s been no shortage of controversy.

Google began 2024 fighting on all fronts amid criticism of its Gemini models after image generation tools sparked a frenzy online. The firm’s decision to rebrand Bard and Duet AI as ‘Gemini’ tools as part of a soft relaunch for its AI range also raised questions about exactly how Google plans to package its AI offerings in the year ahead. 

Google isn't the only one of Apple’s competitors mired in drama linked to its AI ambitions. Microsoft, the leading figure in the generative AI boom at present, has been subject to intense regulatory scrutiny over its ties to OpenAI, which itself was plunged into absolute chaos in November 2023 during Sam Altman’s ousting and eventual return. 

Apple’s quiet, measured, and deliberate approach to AI development comes in stark contrast to the cacophony everyone is contending with at present. Maybe the strong, silent type is exactly what the industry needs. 

A vulnerability etched into the design of Apple M-series chips has been uncovered by researchers which could allow attackers to extract encryption secret keys when performing cryptographic operations.

Six academic researchers at institutions across the US authored a paper outlining a vulnerability they dubbed ‘GoFetch’, which leaks cryptographic data from the CPU cache that hackers can use to piece together a cryptographic key.

“GoFetch is a microarchitectural side-channel attack that can extract secret keys from constant-time cryptographic implementations via data memory-dependent prefetchers (DMPs).” stated a blog published by the authors.

GoFetch relies on exploiting a relatively new microarchitectural design feature only found on Apple M-series chips and Intel’s Raptor Lake microarchitecture intended to reduce memory-access latency – a common CPU bottleneck.

DMPs proactively load data into the CPU cache before it is directly required, helping to reduce latency between the main memory and CPU.

This technology is vulnerable to cache side-channel attacks which “observe the side effects of the victim program’s secret-dependent accesses to the processor cache”, according to the paper.

During the prefetching process, the DMP must make a series of predictions on what data will be required, based on previous access patterns, and attackers can exploit this side channel to steal information.

A popular workaround neutralizing this threat is constant-time programming, which standardizes the execution time for operations regardless of the size of the input by ensuring the data has no secret-dependent memory accesses.

The new paper from Chen et al. demonstrates how DMPs often compromise the security of constant-time programming by mixing up memory content with pointer values that are used to direct the DMP to load other data. 

“We show that even if a victim correctly separates data from addresses by following the constant-time paradigm, the DMP will generate secret-dependent memory access on the victim's behalf, resulting in variable-time code susceptible to our key-extraction attacks:”, Chen et al explained.

Applications using the GoFetch attack can manipulate data to look like a pointer value, which the DMP treats as an address and brings the data from this location into the cache, which is then visible and leaked over cache side channels.

The vulnerability can be exploited when the cryptographic operation being targeted is running on the same CPU cluster as the malicious application.

The authors stated they will release proof-of-concept code demonstrating GoFetch’s attack path soon.

Flaw in Apple M-series chips is ‘unpatchable’

This vulnerability cannot be patched directly as it stems from the microarchitectural design of the silicon itself, the paper stated.

Notably, Intel’s Raptor Lake CPU architecture doesn’t share this vulnerability with its M-series counterparts, despite sharing the same prefetcher as Apple’s chips.

This shows that the vulnerability can be addressed by altering the silicon, but this will only be available for future Apple M-series architectures, where the CPU architecture will need to be redesigned.

As a result, current M-series chips exposed to the vulnerability cannot be patched in the silicon, and businesses using these devices can only try to mitigate the potential damage a successful exploit could incur using third-party software.

But integrating extra layers of protection into third-party cryptographic software will take a significant toll on encryption and decryption performance, leaving developers with a difficult choice between efficiency and security.

At the time of writing, Apple has not published any release dates for an official fix.

Apple is reportedly working on the development of an AI code generation tool as the firm teases a flurry of upcoming generative AI features and tools over the next several months. 

Reports from Bloomberg suggest the company has been working on the tool for the last year as part of a major update to Xcode, the company’s existing programming software used by developers.

The new AI programming tool will reportedly function similar to GitHub’s Copilot tool, which enables users to generate code based on natural language prompts and convert code across multiple programming languages.

GitHub Copilot has been immensely popular among developers since its launch. The AI coding assistant now boasts over one million users, and Microsoft has been keen to roll out new capabilities for the platform.

Recent updates to GitHub Copilot included the launch of new live chat features that can support programmers in real-time, offering advice on code curation.

This new Xcode AI tool will be able to automatically create and finish lines of code, which Apple hopes will enable developers to rapidly accelerate app creation.

Testing of the new functionality has begun internally, Bloomberg reported, and the firm could be preparing to launch the feature to third-party software developers “as early as next year”.

Apple picks up pace in the generative AI race

The move from Apple comes amid a period of intense speculation over its generative AI plans. The tech giant has outlined plans to integrate the technology within Siri and a range of other applications.

But so far, Apple’s foray into the AI race has been subdued compared to other major industry stakeholders such as Microsoft, or Google.

That’s not to say that Apple isn’t making significant strides in this regard, however.

In December, the company released a raft of open source tools as part of a move to provide greater machine learning capabilities on Apple Silicon.

Known as ‘Apple MLX’, the machine learning framework was developed by the firm’s ML research division, and allows users to build AI tools using its in-house silicon.

The MLX framework is available through open source libraries such as PyPI and GitHub, Apple confirmed at the time.

Prior to the launch of Apple MLX, speculation over the extent of Apple’s generative AI ambitions had been growing for some time. Reports from July 2023 suggested that the company was developing its own internal large language model (LLM) known as ‘Apple GPT’.

Details on this model have yet to be confirmed by the tech giant.

CEO Tim Cook has also been reserved on the matter, but in an August interview with CNBC he revealed the firm had plans to ramp up the integration of generative AI capabilities across its product offerings.

As a result of Covid, we have seen a dramatic rise in transformative flexibility in the workplace. And while the 9-5 office is still a mainstay, it’s far more likely that personnel will be able to choose flexible hours and remote or hybrid working.

As other elements of work have become more flexible, the ability to offer employee choice on work devices is an increasingly useful competitive advantage. However, whilst firms may be open to the idea, successful implementation remains unexpectedly difficult for them.

Ultimately, channel partners can play an important role in making employee choice possible and ensuring a smooth transition.

What is employee choice and why is it valuable? 

Employee choice generally means allowing employees to use their preferred technology, that they are most familiar and comfortable with and therefore most productive using. 

Many surveys in recent years have indicated that employees favor Mac over Windows PC if given the choice. iPhones have become commonplace as corporate phones, and the rise of Mac adoption is also evident in the modern workplace, no longer restricted to executive roles or specialist functions like design.

It’s been an increasingly prominent trend in recent years, but really came under the spotlight when IBM announced it would be implementing the policy in 2015. The uptake of the Mac@IBM program demonstrated lower cost of ownership vs Windows PCs, plus a notable improvement in employee engagement scores and IBM’s ability to attract and retain talent.

Employee choice is especially valuable when it comes to attracting and retaining talent, particularly younger workers. We’re seeing more students coming out of university only having used iPhones, iPads and Macs during their studies. Then they walk into an employer that hands them an unfamiliar PC with a Windows OS, and no other option.

If an organization is promoting a policy of letting employees use whatever they prefer, it’s very likely to attract top talent away from less flexible rivals. Research conducted by Jamf found that 70% of workers would be more likely to join a company that offered a choice in tech, and 75% would be more likely to stay.

With more organizations realizing the competitive advantages of employee choice, there’s a growing opportunity for channel partners. Those with the right portfolio and knowledge can take on long-term projects with recurring revenue and build their reputations as trusted advisers.

What are the barriers to implementing employee choice? 

So why is employee choice such a challenge for many organizations? There are both cultural and technical issues to tackle. 

On the cultural side, there is often a misapprehension that Apple products are largely consumer tech and aren’t suitable for the corporate environment. This is an outdated view that doesn’t reflect the state of modern Apple devices. Indeed, our research found that workers able to choose Apple often saw benefits in productivity and creativity.

Alongside business and IT heads reluctant to sign off on devices they mistakenly perceive as cost prohibitive or unsupportable, bringing in Apple endpoints also introduces some technical hurdles for those only used to dealing with Windows. We see this as a great opportunity for channel partners to provide the technical support, knowledge and services to help.

One challenge that still exists in some organizations is the use of legacy, or even homegrown applications. This can hamper the use of modern or non-Windows Operating Systems and also represent a security threat. 

So, partners have an opportunity to work with customers to modernize the applications they use, as a side benefit of introducing employee choice within an organization.

This process of kickstarting modernization can also usher in a new wave of software purchasing along with new hardware.

More important than specific apps, Windows-centric IT and security teams may also not be familiar with best practices for provisioning Apple devices to meet the same corporate policies as the Windows devices they provide to employees.

The growing opportunity for channel partners 

This is where the channel has a huge opportunity ahead of it. Being able to solve the problems around employee choice provides a powerful ‘in’ for engaging with new prospects and furthering the relationships with existing customers. 

The prospect is perhaps not as compelling for partners that are more focused on hardware numbers, especially if they feel Windows provides more volume. But for those that earn their crust by selling their time rather than moving product, employee choice offers a long running professional services opportunity.

Partners can fulfill that all-important trusted advisor role by providing guidance on how a business can best pursue a more flexible approach. What pain points are around Windows-centric hardware and software? What policies will need to be altered or created? What tools will be needed to manage everything?

It’s also important to take a step back for a strategic view. What does the company want to achieve with employee choice? What involvement is required from other departments like HR as well as the expected IT and security teams?

A good partner will not only help answer these questions but have a portfolio with the right products and services at the ready too. The best approach is to bring in solutions that operate in a similar manner to familiar tools like 365 that help them perform identity, management and security in their Windows stack.

Successfully implementing employee choice will mean an organization has more opportunity and incentive to grow its tech stack. This means two bites of the cherry for channel partners, offering hardware and software sales, as they can connect customers with more endpoints and the software licenses to go with them. 

Further down the line it also means more tech refreshes, all leading to a natural ARR growth curve.

By helping their customers stand out from their competition, partners will themselves standout in the channel.

A GPU memory vulnerability dubbed ‘LeftoverLocals’ could expose LLM responses to hackers through leftover local memory, researchers have warned. 

Apple, Qualcomm, AMD, and Imagination are among the big name GPU vendors named as vulnerable according to research posted on the Trail of Bits blog. The GPUs have all been vulnerable to varying extents from as far back as September 2023, when researchers first began their investigation.

Researchers were able to build a proof of concept (PoC) of the potential attack, citing an attached video in which they listen in to another user’s interactive LLM session by recovering a GPUs local memory.

From their tests, the researchers concluded that ‘LeftoverLocals’ could leak around 5.5mb per GPU invocation on an AMD Radeon RX 7900 XT, enough data to reconstruct an LLM with worryingly high precision.

With GPUs used extensively to support the high performance requirements of AI inferencing, this news is likely to raise serious concerns among enterprise AI users.

Eleanor Watson, IEEE member and AI ethics engineer at Singularity university, told ITPro that although this particular vulnerability would require physical access to a GPU, the research highlights serious data protection risks.

“Whilst this particular type of exploit requires direct access to the GPU and its memory, it’s indicative of the challenges of keeping our interactions with AI systems private,” she said.

“I expect that further vulnerabilities will be uncovered which are broadly applicable to a wide range of LLM systems, leading to a widespread doxxing of interactions and generations, and associated embarrassment,” she added.

Vendors have acted swiftly on GPU memory vulnerability  

To defend against the vulnerability, GPUs need a built in system to clear local memory between kernel calls. Some GPU vendors, including Nvidia and Intel, are already doing this, whereas others need to keep pace. 

In response to the news of ‘LeftoverLocals’, AMD stated plans to create a new mode that “prevents processes from running in parallel on the GPU and clears local memory between processes on supported products”, with an expected rollout of said mode by March 2024.

“This mode would be designed to be set by an administrator and not enabled by default,” the firm said in an advisory. “Supporting documentation for the new mode, along with details of how to update AMD products, will be provided in a future update to this security notice.”

Apple responded to Trail of Bits but did not issue specific details of their patch, while Qualcomm issued a partial fix and Imagination a full patch in December.

A failure to patch this vulnerability could see attackers targeting a variety of GPU applications and LLM sessions, including those within privacy-sensitive domains.

Open source LLMs, however, are still the main target. Despite their ability to be rigorously audited, their reliance on closed-source GPUs means they are particularly susceptible to this attack.

“A lot of security hardening will need to be done on AI systems to make them more resistant to these kinds of vulnerabilities”, Watson said.

“This is a necessary component of the ongoing professionalization of AI, along with mastering challenges such as confabulation/hallucination in models”, she added.

Apple is set to make fundamental changes to its App Store in the EU to comply with the European Commission’s Digital Markets Act (DMA), but some experts are worried these changes could have serious negative security implications.

The DMA will introduce a mandate on Apple, and other designated ‘gatekeepers’ to start letting users install third-party app stores and sideload apps onto their devices, aiming to increase competition on online platforms. 

Security specialists are concerned this decision will open up another attack vector for threat actors, however, and potentially result in a flood of dangerous applications being downloaded.

Jamie Moles, senior technical manager at ExtraHop, told ITPro the DMA could have unintended consequences for security of online platforms by removing the operator’s ability to freely moderate the software being distributed.

“It [the DMA] would increase the overall risks to the platform because it creates an entirely new attack vector,” he said. 

“The requirements for an app to appear on the app store are, in part, designed to protect the security of the system and comply with Apple’s own privacy policy. If you remove the ability to directly audit the programs being distributed to the platform, then it is inevitable that there will be more security risks.”

For example, Moles highlighted a recent example of security flaws in the Epic Games Store, which could be accessible on iOS under the new DMA stipulations.

“The Epic Games Store (EGS), a platform that could be added to iOS if the courts rule in Epic Games’s favor, had issues with privacy upon release. EGS was deriving contacts from other games stores by directly accessing files stored locally by a competing app store, Steam, rather than using a more secure API link”, Moles commented.

“Epic Games apologized but has not changed the method by which it syncs contacts with Steam contacts. This is a security risk created by a company with good intentions, so the potential downsides for a company with bad intentions are huge.”

How will Apple respond to these changes?

Moles said he expects the changes to the Apple App Store to be limited to the EU region, as opposed to the regulation prompting a universal policy shift from the tech giant like it did with the mandated USB-C connectors for iPhones.

“Another app store is a software change, and Apple already has different software configurations for different regions. In October, Apple required an Internet Content Provider license for apps to appear on the Chinese version of the app store, so the precedent is there for each region to have its app store configured for different regulatory environments.”

Apple could introduce a series of permission layers on iOS to protect the majority of its users from inadvertently accessing unaudited marketplaces or sideloading malicious applications, Moles suggested. 

Another measure Apple could use to mitigate the risks associated with this shift could be to force unverified apps to be opened in a dedicated software environment to minimize its access to the device's files and other applications.

Moles added that Apple could implement a similar approach to its policy on aftermarket hardware components, such as batteries or screens.

“Imagine when downloading or booting one of these applications, several layers of warning notifications for each permission being granted to the software will appear on screen”,  he explained.

“The implementation would be similar to those in place for apps already, but likely in more explicit terms. Installing and sideloading could result in voiding warranty for devices, something Apple already does with aftermarket batteries and screens.”

Balance is needed to prevent monopolies and limit security risks

The DMA was first tabled in 2022 to address the significant advantage big tech companies have over smaller competitors in the industry by limiting the control these companies have over how business is conducted on their platforms.

The regulation is targeted at preventing gatekeepers from using their privileged position as platform operators to promote their own products or services above those of their competitors. 

On the subject of whether this act will produce a net benefit for consumers in the long run, Moles said, overall, the EU’s tech regulation in recent years has been warranted and in the interest of business and consumers alike, but it needs to be careful it isn’t creating more problems than it solves with the DMA.

“Most of the EU regulation in the tech space has been a good thing. The USB-C requirements are, by and large, to the benefit of consumers and GDPR, and as annoying as cookie notices can be, they set important protections on user data” Moles told ITPro.

“However, while well intentioned, this particular change exposes users to far more risk with less obvious reward. Preventing large tech monopolies is essentially a good idea, but one that has to balance the challenges of creating and operating a platform of iOS’s size.”

European tech companies and industry groups have demanded better communication from tech “gatekeepers” ahead of the introduction of the EU’s Digital Markets Act (DMA).

An open letter from tech companies including VPN firm Proton and search engine Qwant, as well as industry bodies including the European Tech Alliance, said that the thousands of businesses they represent were worried about a lack of "effective" engagement from the companies that have been designated gatekeepers under the DMA.

“Gatekeepers have either failed to engage in a dialogue with third parties or have presented solutions failing short of compliance with the DMA,” the letter said.

“Businesses and consumers are largely kept in the dark as to what is going to happen after 7 March 2024,” the letter said.

In September 2023, the European Commission said that Google’s parent company Alphabet, along with Amazon, Apple, ByteDance, Meta, and Microsoft were considered ‘gatekeepers’ under the DMA.

Under the terms of the legislation, the European Commission can designate companies as gatekeepers if they provide a key ‘gateway’ between businesses and consumers through their core platform services.

In this case, the EC said it counted 22 core platform services in total across the six gatekeepers in areas including search, social networks, browsers, operating systems, and video sharing.

The new rules establish new obligations for gatekeepers, which have six months to ensure compliance for each of their core platforms. That means they have until 7 March to get ready.

How the Digital Markets Act will work

For example, in some situations, they may have to allow third parties to interoperate with their own services, enable business users to access the data they generate by using the gatekeeper’s platform, and provide companies advertising on their platform with tools to carry out their own independent verification of their ads.

Gatekeepers will also be forced to allow businesses to promote their offers and make deals with customers outside the gatekeeper’s platform.

According to the EC, gatekeeper platforms can no longer treat their own services and products more favorably in ranking, or prevent consumers from linking up to businesses outside their platforms.

They also can’t stop users from uninstalling any pre-installed software or app, or track end users outside of the gatekeepers' core platform service for advertising without their explicit consent.

The consequences of non-compliance can be painful for these gatekeepers, including fines of up to 10% of the company’s total worldwide annual turnover, or up to 20% in the event of repeated infringements.

In their letter, the European tech firms and industry groups said the DMA will affect how millions of consumers and business users interact with social networks, app stores, online shopping, video sharing services, and more.

“It will constitute a paradigm shift in digital markets, with one fundamental objective: the creation of fair and contestable digital markets in Europe,” the letter reads. “It should put an end to anticompetitive practices that led to higher prices for consumers and slowed innovation in Europe.”

The launch of the Digital Markets Act in March 2024 will mark the “beginning of the new era”, the consortium said, and one that will improve competition for smaller businesses across the union.

“It would be regrettable if that new era began with a false start, which will happen if the gatekeepers do not constructively engage with third parties, including business users and consumer associations, before 7 March on how they intend to comply with the DMA,” it said.

The letter said the gatekeepers should engage “as soon as possible” with businesses and others in a constructive dialogue and make "swift progress on their proposed compliance solutions".