The trillion-dollar question: Who pays when the industry’s AI bill comes due?

Over the past two and a half years, hyperscalers and other large IT vendors have spent over a trillion dollars to ramp up their AI capabilities, with investments far outpacing profits in the short term.

A year ago, VC firm Sequoia Capital projected AI investments would outpace AI-related revenue by $600 billion at the end of 2024. While AI-driven companies like ServiceNow and Salesforce are reporting significant revenue growth, AI spending is projected to rise this year to $644 billion, up more than 76% from 2024, according to Gartner. AI investment in 2024 was up a whopping 337% from 2023.

The trillion-dollar question about all this spending then becomes: Who will be stuck with the bill? Using a safe assumption that AI vendors and companies like Microsoft, OpenAI, Google, and IBM aren’t going to eat the costs, CIOs might believe they are on the hook.

AI market observers don’t all agree about the timeframe or the methods of how AI vendors will recoup their investments, but many warn CIOs that a reckoning is coming. The cost of popular AI tools will eventually increase, according to some, while others see CIOs soon paying for many new products. Another outlook sees an AI bust coming, with investors taking the hit.

Longtime IT VC Nick Davidov, co-founder of the Davidovs Venture Collective, believes AI costs will go down over the long term as the technology matures, and he notes that hundreds of billions of dollars in investment per year is a small fraction of the $5.6 trillion in worldwide IT spending Gartner projects for 2025.

Still, Davidov sees IT budgets shifting toward more spending on AI and cuts in other areas, including traditional SaaS products. In addition, many companies will cut their workforces through AI efficiencies, with some of the savings paying for new AI tools.

He also predicts IT budgets will grow as AI generates new revenue streams at customer organizations.

“The CIO is going to be very, very busy for the next three, four years, and that’s going to be the biggest impact,” he says. “All of a sudden, businesspeople are starting to figure out that they can save a ton of money with AI, or they can enable their best performers to do the actual job.”

Davidov doesn’t see workforce cuts matching AI productivity increases, even though some job cuts may be coming.

“Just because people suddenly realize that if they can increase their efficiency in certain processes by like 30% or 40% it doesn’t necessarily mean that they need to cut 40% of the workforce in that process, like they would in a vacuum,” he says. “They have competition that is also starting to use AI. So now they’re like, ‘I’m not going to lay off all the people today. I have to use AI tools to be more efficient, because my competitors are doing the same thing.’”

A new dot-com boom

Other observers compared current AI investments to the dot-com boom of the late 1990s.

“All these VCs are racing to put their hands on the next unicorns in this emerging technology,” says Arie Brish, a business professor at St. Edwards University. “Obviously, 95% will fail, but 5% will be the next Google, Amazon, or Meta. That’s life in the venture world.”

Customers will pay for at least some of the investment, he predicts. “In some cases, customers are already paying because there will be added features in products they already pay subscriptions for,” he says. “In cases of totally new products that don’t exist today, it will be on a case-by-case basis, and each organization or CIO will have to do a cost-vs.-benefit analysis of these products.”

Rick Bentley, founder of AI surveillance firm Cloudastructure, had a front-row seat for the dot-com boom and bust, and he sees a repeat cycle happening. “There were winners and losers,” he says. “You can only lose billions of dollars per year for so long. Google and Meta are still around and doing great. What about countless others?”

If an AI bust is coming, customers won’t be on the hook for all those investments, Bentley says.

“It’s more that the casino is open, drinks are flowing, people are gambling wildly, and the staff is getting tipped well,” he adds. “We, the customers, are the staff raking in tips. Whether the gamblers win or lose, the valet gets to keep the $100 he got for parking the Lambo out front, and we get to use free ChatGPT accounts.”

New pricing models

Other observers see AI vendors’ customers footing a large portion of the bill. CIOs are likely to face higher subscription fees, usage-based pricing, or premium charges for advanced features, says Firdaus Bhathena, CTO at FinTech firm Fidelity Information Systems.

“The costs of building out AI infrastructure will ultimately fall to enterprise users, and for CIOs, it’s only a question of when,” he says. “While hyperscalers and AI vendors are currently shouldering much of the expense to drive adoption, we expect to see pricing models evolve.”

Bhathena advises CIOs to look beyond headline pricing because hidden costs, particularly around integrating AI with existing legacy systems, can quickly escalate. Organizations using AI will also need to invest in upskilling employees and be ready to navigate increasingly complex vendor ecosystems.

“Now is the time for organizations to audit their vendor agreements, ensure contract flexibility, and prepare for potential cost increases as the full financial impact of AI adoption becomes clearer,” he says.

JB Baker, vice president of products at next-generation storage and memory vendor ScaleFlux, comes at the AI investment conversation from a different angle. He argues that legacy hardware and infrastructure are driving up the cost of AI development, with GPU advancements outpacing other hardware.

Nevertheless, he believes AI vendors will find new ways to drive revenue. “Eventually, it’s not going to be so much an increase in cost of services that you consumed already, but it’s new services that you didn’t know you needed,” he says.

Baker advises CIOs to be careful about their purchases of AI products and services and tie new deployments to business needs. “If you’re just buying AI to keep up, you may spend a lot of money and not get an ROI out of it,” he adds. “Define ways that it’s going to help your business. How is this going to improve our bottom line, our revenue?”