Looking for workarounds: Why some employees turn to shadow AI when outages hit – and how IT leaders can stop it
IT leaders need to accept that staff will try to find workarounds. It’s up to them to set clear ground rules
Waking up to find Microsoft 365 was down might’ve been a welcome reprieve for some employees, but for others it created a dangerous predicament.
With deadlines approaching and being unable to access the AI tools they rely on, research shows that many turned to alternative options – and in doing so they placed themselves and their business at risk.
Analysis from CultureAI found that in the wake of the recent Microsoft 365 outage, one-in-eleven Copilot users switched to another AI tool as a workaround during the incident.
Notably, nearly half (40%) turned to free accounts or tools that didn’t require an account at all. Adding insult to injury, one-in-three of those entered personal or sensitive business data into said tools.
Similar research from Harmonic Security also identified a 120% spike in Copilot users accessing free AI tools during the outage.
Speaking to ITPro, Muhammad Yahya Patel, vCISO and cybersecurity advisor for EMEA at Huntress, said outages can create the perfect storm when it comes to “shadow AI”.
Left in the dark, users scramble for alternative options to help support them in their daily workflow.
Sign up today and you will receive a free copy of our Future Focus 2026 report - the leading resource for IT decision-maker insight on priorities and investment areas in AI, security and more.
“This research captures something the industry hasn’t been talking about loudly enough. Outages don’t just disrupt productivity,” he told ITPro.
“They create the exact conditions under which people make poor security decisions: time pressure, frustration, and a workaround that feels harmless in the moment.”
Emanuela Zaccone, principal AI security strategist at Sysdig, echoed Patel’s comments, albeit noting that instances such as these merely expose existing problems with unauthorized AI use.
“When an organization’s AI governance plan assumes employees will rely on one sanctioned tool to always be available, an outage will show them what happens when it isn’t. People have work to do and deadlines to meet, so they will reach for whatever tools are closest,” she told ITPro.
“For employees, the move isn’t malicious. It’s a continuity problem, and IT departments have planned for those for years. The difference here is that AI has become load-bearing for many people’s daily work,” Zaccone added.
“When something is load-bearing, you have to make it resilient. People should still be able to get work done with the workflows they’ve come to expect regardless of whether one tool goes away. AI shouldn’t be a single point of failure that stops operations.”
The perils of shadow AI
Shadow AI is by no means a new trend. Indeed, it’s a topic that’s been front and center for IT and security teams since the advent of the generative AI race. Using these tools causes significant problems for enterprises, largely because IT teams cannot track what tools employees are using.
This creates blind spots from a security perspective and adds new layers of complexity – and it’s a trend that shows no signs of slowing down. Research from Gartner in November 2025 projected that around 40% of organizations will experience shadow AI-related breaches by 2030.
More recently, the UK’s National Cyber Security Centre (NCSC) issued a call to action for enterprises to implement more robust governance measures. The security agency warned that unauthorized AI tools can introduce “new cybersecurity risks that can be hard to identify.”
“You cannot manage what you do not know,” the NCSC said in an advisory.
How IT leaders can tackle shadow AI
For IT and security leaders, tackling unauthorized AI use starts with improving visibility and accountability. Zaccone said teams need to audit what’s being used, where it’s being used, and who has access to the technology.
“The first step toward better AI governance is visibility. That means taking an inventory of what’s actually running on endpoints,” she said. “From there, I’d encourage leaders to name an approved fallback tool before an outage, so nobody has to choose another one under pressure.”
Clear guidance on shadow AI is vital, Zaccone added, enabling IT leaders to embed a security-conscious culture among staff. Fallback options are also advised to ensure continuity in the event that one particular tool goes down.
“You can build that guidance into the incident communications plan, so staff hear the same advice on a normal day and on the day something breaks,” Zaccone noted. “You also may find that some employees prefer the option to use multiple tools, some of which may perform better for certain tasks anyway.”
Michael Marriott, VP at Harmonic Security, told ITPro that acknowledging that staff will seek workarounds is also a solid starting point. Thereafter, communicating expectations and best practices becomes easier, he noted.
“Start from the assumption that workarounds will happen and establish visibility across both corporate and personal accounts,” Marriott commented.
“Rather than blocking tools, apply controls based on the content being shared so that an employee analyzing a public article in a free ChatGPT account proceeds without interruption, while one pasting in a customer contract is blocked inline with an explanation. This permits some AI use during an outage, even if the high-level confidential work has to wait until an outage resolves itself.”
In a similar vein to Zaccone’s advice, Patel told ITPro that having an approved list of alternative tools and fallback options could prove critical for organizations in the event of a future outage.
“Pre-approve a small set of alternative AI tools with appropriate data handling agreements, communicate them clearly, and make them easy to access so users don’t have to improvise,” he said.
If disaster strikes and users are once again left in the dark, IT leaders need to push guidance to staff to remind them “what’s acceptable and what isn’t”.
FOLLOW US ON SOCIAL MEDIA
Follow ITPro on Google News and add us as a preferred source to keep tabs on all our latest news, analysis, views, and reviews.
You can also follow ITPro on LinkedIn, X, Facebook, and BlueSky.
Ross Kelly is ITPro's News & Analysis Editor, responsible for leading the brand's news output and in-depth reporting on the latest stories from across the business technology landscape. Ross was previously a Staff Writer, during which time he developed a keen interest in cyber security, business leadership, and emerging technologies.
He graduated from Edinburgh Napier University in 2016 with a BA (Hons) in Journalism, and joined ITPro in 2022 after four years working in technology conference research.
For news pitches, you can contact Ross at ross.kelly@futurenet.com, or on Twitter and LinkedIn.
-
What is the best approach to open standards in the age of AI?Podcast QNX's engineer Sheridan Ethier joins Jane and Bobby to discuss how the arrival of AI has exacerbated open source standards
-
Oppo Pad 5 Pro reviewReview It's not quite the creative powerhouse Oppo suggests, but a very clever AI-filled device, nonetheless
-
How can IT leaders tackle shadow agentic AI threats?Insights Unchecked AI agents exploit employee access to alter critical systems. Here's how IT leaders can regain control
-
OpenAI says its new cyber-focused AI model found thousands of Linux kernel vulnerabilitiesNews The new cyber-focused AI model boasts improved bug hunting capabilities compared to its predecessor
-
88% of UK firms have had an “agentic security incident” in the last year – ServiceNow warns more will comeNews Concerns are growing over the scale of adoption and an inability to match “machine-speed” capabilities
-
‘Rising threats and under-resourcing for cybersecurity is taking a toll on the people tasked with managing it’: Cyber teams are being pushed to breaking point – and AI is doing little to alleviate strainNews An increasingly perilous threat landscape and lack of funding means cyber teams are being stretched too thin
-
'You cannot manage what you do not know': The NCSC is calling for a crackdown on shadow AINews Organizations are urged to identify shadow AI use and tighten up security procedures
-
'It delivers world-class performance at 50 percent of the cost of leading models': Microsoft unveils cut-price AI for security with latest in-house model launchNews Pairing the MAI security model with GPT-5.4 gives benchmark leading results at half the cost, according to the tech giant
-
The case for the channel in an AI-driven security marketIndustry Insights AI won't replace channel partners; SMB cybersecurity still relies on trust
-
Cisco just launched two cyber-focused small language models: Antares-350M and Antares-1B aim to supercharge codebase analysis – and they run at a “fraction of the compute expense” of popular frontier modelsNews The Antares models unveiled by Cisco aim to cut costs in codebase analysis