Google shakes off tariff concerns to push on with $75 billion AI spending plans – but analysts warn rising infrastructure costs will send cloud prices sky high
Industry analysts warn Google and competitors could face huge cost increases – and that will ultimately trickle down to customers
Google CEO Sundar Pichai has confirmed the company will still spend $75 billion on building out data centers despite economic concerns in the wake of US tariffs.
While many were paused or reduced yesterday, heavy tariffs still remain on China, potentially raising the cost of infrastructure projects deemed necessary to enable the development and rollout of AI.
Speaking at the Google Cloud Next conference, Pichai took to the stage to confirm that the $75 billion investment was still planned for this year.
"The opportunity with AI is as big as it gets," he said. "That's why we are investing in the full stack of AI innovation, starting with infrastructure that powers it all. We are making big investments now, and for the future."
"This investment will be directed towards our servers and data centers, which includes power our AI compute and cloud business," he added. "So this will greatly benefit our customers, like all of you."
He added that the infrastructure spend will benefit existing services like Search and Google Workspace, but will also be used for training its Gemini model.
That $75 billion was originally announced by Google's parent company Alphabet in February as part of quarterly results reporting. Last year, Google spent $52.5 billion on capital expenditures, marking a significant increase.
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.
Similar infrastructure investment is planned by rivals. Microsoft said it would spend $80 billion on AI equipment this year, while Meta announced spending of $65 billion.
That comes alongside huge data center investment as part of Project Stargate, a $500 billion plan to build AI infrastructure in the US that is backed — but not paid for — by the US government.
Enterprises could feel the pinch
The high cost of infrastructure required to support AI development was already causing concerns. Last year, Microsoft said that it could take 15 years for the investment to pay off and investors at a host of major firms have voiced worries over spiraling capital expenditures.
With the arrival of tariffs and a looming trade war, these costs could surge even higher, according to industry analysts.
Mark Moccia, VP research director at Forrester, warned IT infrastructure in particular will experience “significant price increases” as major manufacturing nations face high tariff rates.
"The rising costs could balloon budgets and force CIOs to delay or prioritize the most important projects,” Moccia said. “CIOs and other tech leaders will need to proactively analyze costs, diversify sourcing, optimize inventory, and prioritize the projects that don’t sacrifice critical AI ambitions."
Similarly, Forrester principal analyst Lee Sustar said ambitious infrastructure plans and multi-billion dollar data center build outs will become “significantly more expensive” due to price increases on building materials.
Ultimately, he said, this could lead to rising cloud costs as major providers compensate for challenging economic conditions.
"At the same time, the demand for cloud services — especially pricey AI offerings — will drop at least in the near term due to uncertainty over the wider economy," Sustar said.
"Cloud providers will face pressure to pull back on big investments and pass costs to customers with price increases."
MORE FROM ITPRO
- Google CEO says more than 25% of the company's code is now AI-generated – is this the future of software development?
- Keep up to date with all the news and announcements from Google Cloud Next 2025
- Foreign AI model launches may have improved trust in US AI developers, says Mandiant CTO
Freelance journalist Nicole Kobie first started writing for ITPro in 2007, with bylines in New Scientist, Wired, PC Pro and many more.
Nicole the author of a book about the history of technology, The Long History of the Future.
-
Post-cloud strategy: Architecting the next enterprise stackAs enterprises rethink their dependence on hyperscale, hybrid architectures are emerging as the new foundation for resilient, AI-ready infrastructure
-
Anthropic just launched Claude Fable 5, its first Mythos-class AI modelNews The launch of Claude Fable 5 marks the first public release of a Mythos-class AI model
-
Google once again ramps up funding for AnthropicNews Google is once again backing Anthropic with a significant funding boost
-
Google Cloud announces eighth-generation TPUs, boasting AI training and inference leapsNews Alongside chip improvements, Google Cloud has overhauled its data center fabric and file systems to support enterprise-scale AI deployment
-
Has another OpenAI Stargate project hit the rocks?News Microsoft will be renting capacity in Norwegian data center originally intended for OpenAI
-
Anthropic pens multi-gigawatt TPU deal with Google and Broadcom as Claude demand picks upNews The frontier model maker is looking to greatly expand its compute infrastructure to meet surging customer demand
-
Google talks up new Texas data center that uses minimal water – except for 'critical campus operations' like kitchensNews New data center plans look to capitalize on air cooling and clean energy generation
-
Going all-in on digital sovereigntyITPro Podcast Geopolitical uncertainty is intensifying public and private sector focus on true sovereign workloads
-
Sovereign infrastructure spend to triple in Europe as fifth of workloads stay localNews Gartner says global spending on sovereign cloud infrastructure will climb 35% over the next year
-
Google drops $4.75bn on data center and energy firm IntersectNews The investment marks the latest move from Google to boost its infrastructure sustainability credentials
