The $63.4 Billion Innovation Drain: Study Finds Unexpected Tech Costs Force Cuts into AI Investment and Hiring

New economic research reveals how unpredictable infrastructure costs are constraining innovation, with potential savings that could support more than 111,000 jobs and generate $20.4 billion in long-term growthAI/ML infrastructure overspend is the most of any technology category at 37% ($23.4bn) of the total, and nearly twice as hard to forecast84% of organizations made at least one cost-driven trade-off over the past year

Businesses across eight major economies incur $63.4 billion annually in technology infrastructure overspend attributable to opaque, unpredictable or difficult-to-forecast provider charges, forcing companies to scale back investments in artificial intelligence, hiring and new product development, according to new economic research from the Centre for Economics and Business Research (CEBR), commissioned by Wasabi Technologies. The overspend reflects infrastructure provider charges businesses could not forecast, rather than their own planned usage.

AI and machine learning infrastructure accounted for $23.4 billion, or 37%, of the total overspend, more than any other technology category. It is also the category businesses find hardest to forecast. AI infrastructure costs are unpredictable for 22% of businesses, nearly double any other layer in the technology stack, where the range is between 8% and 12%. The findings expose a growing financial challenge for companies investing in AI. The infrastructure needed to support new technology is consuming budgets intended to advance it.

“Our research uncovered the hidden cost of innovation for some of the world’s largest economies and found the economic cost of unpredictable pricing is not simply that firms pay more,” commented Owen Good, head of economics advisory at CEBR. “This means hiring and investment in new technologies are postponed or reduced as budgets are redirected to cover unexpected charges. The technology generating the most unplanned costs is also among the investments businesses are cutting to pay for them.”

When Infrastructure Costs Rise, Innovation Pays the Price

The findings suggest a widening gap between corporate ambitions for AI and the financial realities of deploying it.

As companies move from experimenting with AI to deploying it at scale, usage-based pricing is adding another layer of cost uncertainty. Among businesses reporting AI and machine learning infrastructure overspend, 38% cited higher-than-expected token, inference or usage-based AI charges as a contributing factor. Those costs ranked alongside unexpected AI platform and model licensing fees (49%) and integration, deployment and model operations expenses (45%) as leading drivers of AI overspend.

The unpredictability translates directly into financial strain. Among businesses experiencing unpredictable infrastructure costs, 73% reported moderate or significant budget pressure. Among those facing that pressure, 36% had already delayed or scaled back AI, automation or data science initiatives.

The consequences extend beyond AI. The study found 84% of organizations made at least one cost-driven trade-off over the past year. The budgets most sacrificed are:

1. Technology and software investment (29%)
2. Training and workforce development (28%)
3. Hiring in technology, data and AI roles (27%)

The research points to a broader economic cost as companies redirect capital away from projects that could support future growth. CEBR modelled what recovering that overspend would be worth to the eight advanced national economies. Recovered across all markets, the overspend would support:

• $21.8bn of new productive investment in R&D, AI and new technology
• $7.6bn of investment into workforce training
• An additional 111,450 jobs
• The above will contribute to a long-run GDP uplift of $20.4bn

Rising Data Movement Costs and Hardware Shortages Limit Escape Routes for Companies

Businesses seeking to control infrastructure spending face a second obstacle: switching providers or expanding capacity can introduce additional costs and delays.

Many faced delays in the last 12 months, with 82% surveyed noting a lack of access to suitable hardware or compute capacity as the cause. For those considering a supplier switch, 89% of businesses incurred additional costs from fragmented data environments, and 84% experienced restricted or delayed access to data needed for a project or decision at least once in the past year.

These constraints are affecting companies' ability to develop and bring new technologies to market. Respondents reported data access challenges had caused them to:

1. Run fewer AI, machine learning or data science experiments (25%)
2. Avoid adopting a new technology or platform (22%)
3. Delay or cancel a product or feature launch (16%)

“Businesses are not short of ambition, but they are blindsided by unpredictable infrastructure costs that prevent them from driving their innovation agendas forward,” said Marty Falaro, president and COO, Wasabi Technologies. “There is a fundamental problem with how technology infrastructure is bought and priced as businesses need to know what it will cost to store, access and move their data before they commit to an AI project, not after the bill arrives. Innovation should take risks, but the infrastructure bill shouldn't.”

Industry Response

“The research reflects what we’re seeing firsthand. AI costs are difficult to predict across licensing, deployment and day-to-day operations, much of which is outside our control," said Andy Fletcher, Vice President, Technology & Digital Product, Liverpool FC. "But how we manage the data behind AI is something we can control. We have decades of footage and more arriving every matchday, so the cost of storing, moving, and accessing that content matters. It’s why we’re taking a closer look at AI Storagemaxxing, making sure data is stored and accessed in the most cost-effective way at each stage of the AI pipeline. The goal is to make better use of the content we already have without adding unnecessary costs or creating more data silos.”

"Vendor lock-in slows down innovation" said Michael Reid, CEO of Megaport. "We added Wasabi to our partnership ecosystem earlier this year so enterprises can significantly reduce costs when deploying the object storage that every organization relies upon. Improved efficiency and cost predictability through a single, global platform allows companies to cost-effectively support workloads like data lakes, analytics, cyber resilience, and backup, delivered at the click of a button around the world."

Reid added that Megaport’s services have two business models: on-demand consumption delivered in 60 seconds, and large, complex compute, network, and storage clusters secured through fully-funded take-or-pay contracts. "All current contracts have contracted data center capacity and are fully funded," Reid said.

“We nearly made a trade-off based on cost. Our storage costs were climbing and the obvious fix was to move older recordings to a cheaper, colder tier, but it would be slow to retrieve data,” said Allan Bettarel, CTO at tl;dv. “For a product where the recording must be ready the moment a meeting ends, that would have meant sacrificing the instant access our customers rely on. We decided to migrate from AWS instead, because with Wasabi we didn’t need to sacrifice cost for performance. Holding our data in their German data centre meant we met data residency and GDPR requirements too. Now I know what we’ll pay for the year, and we’re not charged every time we access our own data.”

Click here to read the full Cost of Innovation Report

[ENDS]

Note to Editors:

Methodology

The research was conducted by Sapio Research in July and August 2026 among 690 senior business and IT decision-makers at small, medium and large organisations across the UK, the US, France, Germany, Italy, Japan, Australia and New Zealand. The margin of error is ±3.7 percentage points at total sample level. Economic modelling, including national overspend estimates and the long-run GDP, investment and employment scenarios, was carried out by Cebr.

About Wasabi Technologies

AI is only as powerful as the data behind it. Wasabi delivers simple, predictable cloud storage that helps organizations build the data infrastructure powering AI, analytics, cyber resilience, media, and cloud-native applications. By giving organizations the freedom to move, access, and activate their data without the hyperscaler tax, Wasabi helps turn data into a competitive advantage. Operating 16 storage regions and supporting organizations doing business in more than 100 countries, Wasabi delivers the global scale, performance, and reliability needed to power the next generation of AI and data-driven innovation. Learn more at wasabi.com.

About CEBR

The Centre for Economics and Business Research (Cebr) is a leading independent economics consultancy based in London. Established in 1992, Cebr has over 30 years of experience providing economic forecasting, research and analysis to businesses, government departments, financial institutions and public organisations worldwide. Its expertise spans macroeconomic forecasting, economic impact assessments, policy analysis and strategic advisory services. Combining rigorous economic modelling with practical insights, Cebr helps organisations make informed decisions, understand economic trends and contribute to public policy debates.

For more information, visit www.cebr.com.

View source version on businesswire.com: https://www.businesswire.com/news/home/20261110600468/en/