Microsoft reports expose AI's cost problem: The tech is more expensive than paying human employees | FortuneSearchSubscribeHomeLatestFortune 500FinanceTechLeadershipLifestyleRankingsMultimediaTrendingnow1Jeff Bezos wants the bottom half of earners to pay zero income tax—he says nurses making just $75K should save $12K a year2Despite a $500 million net worth, Shaq just finished his fourth degree. He warns graduates: 'Your character will take you further than your resume'3Bolt CEO says he let go of his entire HR team for creating problems that didn’t exist: ‘Those problems disappeared when I let them go’ 1Jeff Bezos wants the bottom half of earners to pay zero income tax—he says nurses making just $75K should save $12K a year2Despite a $500 million net worth, Shaq just finished his fourth degree. He warns graduates: 'Your character will take you further than your resume'3Bolt CEO says he let go of his entire HR team for creating problems that didn’t exist: ‘Those problems disappeared when I let them go’ AITechMicrosoft reports are exposing AI’s real cost problem: Using the tech is more expensive than paying human employeesBy Jake AngeloJake AngeloNews FellowDown Arrow Button IconBy Jake AngeloJake AngeloNews FellowDown Arrow Button IconMay 22, 2026, 12:56 PM ETAdd us onMicrosoft CEO Satya NadellaSven Hoppe/picture alliance via Getty ImagesFirms today are pushing employees to use as much AI as possible to squeeze out the technology’s productivity gains. But that pressure is leading to cracks, and those cracks may be irreparable. Recommended Video
Microsoft has reportedly begun canceling most of its direct Claude Code licenses, according to The Verge, instead moving engineers toward using GitHub Copilot CLI. That comes just six months after the firm first opened up access to Claude Code, encouraging thousands of its developers, project managers, designers, and other employees to experiment with coding. The tech became popular fast. Perhaps too popular. The scale at which employees use it is now prompting the firm to reverse course on a tool its own engineers had come to rely on. Canceling Claude Code licenses won’t affect Microsoft’s Foundry deal, which includes investing up to $5 billion in Anthropic and giving Foundry customers access to Claude models, as well as Anthropic’s $30 billion commitment to purchase Azure compute capacity, according to The Verge.
Microsoft isn’t the only company scaling back its internal AI use. Uber’s CTO Praveen Neppalli Naga told The Information in April that the firm had already burnt through its entire 2026 AI coding tools budget in just four months. That comes after the company had actively incentivized adoption through internal leaderboards ranking teams by AI tool usage.
The reports may throw cold water on the bets tech’s biggest firms have placed on the technology. While some cling to the promise of an AI “renaissance” or “revolution,” the cost of adoption is proving a stubborn bottleneck. These developments also suggest that the economics of replacing or augmenting human labor with AI may be more complicated than some early forecasts originally implied. That echoes what Bryan Catanzaro, vice president of applied deep learning at Nvidia, recently said in an interview with Axios. 
“For my team, the cost of compute is far beyond the costs of the employees,” he said.
Anthropic didn’t immediately respond to Fortune’s request for comment. Microsoft didn’t provide a comment. 
An emerging AI paradox: cheaper tokens, bigger bills
Uber and Microsoft aren’t the only firms pushing employees to use as much AI as possible. Like at Uber, a Meta employee crafted a leaderboard, fittingly named “Claudeonomics,” after Anthropic’s AI model, to track which workers are using the most AI. Amazon is pushing its employees to “toxenmaxx,” or use as many AI tokens as possible (the basic building blocks of AI compute).
But with a token-based pricing system, the work gets more expensive with more use and better efficiency. Goldman Sachs recently forecasted that agentic AI could drive a 24-fold increase in token consumption by 2030 as consumers and enterprises adopt AI agents, reaching a staggering 120 quadrillion tokens per month. As businesses turn to AI agents to boost productivity, aggregate costs could rise sharply even if the price of each token falls.
But as consumption increases, the cost of individual AI tokens is expected to fall sharply. A recent report from research firm Gartner found that by 2030, inference on a one-trillion-parameter LLM—in simple terms, a highly sophisticated AI model—will cost AI firms nearly 90% less than it did in 2025. Even so, Gartner predicted that cheaper tokens won’t translate to cheaper enterprise AI because agentic models require far more tokens per task than standard models, increased consumption can outpace falling unit costs, and AI providers won’t fully pass through lower costs to consumers. In turn, inference costs are likely to push higher.
“Chief Product Officers (CPOs) should not confuse the deflation of commodity tokens with the democratization of frontier reasoning,” Gartner senior director analyst Will Sommer warned in a statement.
That reality may complicate the grand plans some firms have for deploying AI agents. Nvidia CEO Jensen Huang recently said he thinks 100 AI agents will one day work alongside every employee at his company. 
Huang is part of a broader wave of CEOs touting an agentic future in which digital workers operate across the enterprise. But if token consumption rises faster than unit costs fall, that future could come with a much heavier bill than executives expect.
In 2001, Fortune first convened the smartest people we know, bringing together CEOs and founders, builders and investors, thinkers and doers. Since then, Fortune Brainstorm Tech has been the place where bold ideas collide. From June 8–10, we will return to Aspen—where it all began—to mark 25 years of Brainstorm. Register now.About the AuthorBy Jake AngeloNews FellowSee full bioRight Arrow Button IconLatest in AIFinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025Most PopularFinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025FinanceLorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniamBy Fortune EditorsOctober 20, 2025Fortune Secondary LogoRankings100 Best CompaniesFortune 500Global 500Fortune 500 EuropeMost Powerful WomenWorld's Most Admired CompaniesSee All RankingsLists CalendarSectionsFinanceFortune CryptoFeaturesLeadershipHealthCommentarySuccessRetailMpwTechLifestyleCEO InitiativeAsiaPoliticsConferencesEuropeNewslettersPersonal FinanceEnvironmentMagazineEducationCustomer SupportFrequently Asked QuestionsCustomer Service PortalPrivacy PolicyTerms Of UseSingle Issues For PurchaseInternational PrintCommercial ServicesAdvertisingFortune Brand StudioFortune AnalyticsFortune ConferencesBusiness DevelopmentGroup SubscriptionsAbout UsAbout UsPress CenterWork At FortuneTerms And ConditionsSite MapRankingsSectionsCustomer SupportCommercial ServicesAbout UsPress CenterWork At FortuneTerms And ConditionsSite MapFacebook iconTwitter iconLinkedIn iconInstagram iconPinterest iconLatest in AIAITransportationHow Grab’s CTO sees the superapp’s push into physical AI and automated driving—and why he uses his competitors’ robots in the officeBy Angelica AngMay 22, 20268 hours agoAIAmerican PoliticsTech billionaires convinced Trump to back off an AI executive order. But much of MAGA favors AI regulationBy Jeremy KahnMay 22, 20268 hours agoAIbooksBarnes & Noble CEO clarifies the bookseller’s stance on AI-written books after refusing to ban them: ‘This is a straightforward rejection of AI books’By Sasha RogelbergMay 22, 202610 hours agoAITechMicrosoft reports are exposing AI’s real cost problem: Using the tech is more expensive than paying human employeesBy Jake AngeloMay 22, 202612 hours agoAIOpenAIThe big questions looming over OpenAI’s trillion-dollar IPOBy Beatrice NolanMay 22, 202612 hours agoSuccessCareersApple’s Steve Wozniak says he cofounded the tech giant after 5 rejections from HP—not to ‘make money.’ For years, his paycheck was just $50By Preston ForeMay 22, 202614 hours agoMost PopularSuccessJeff Bezos wants the bottom half of earners to pay zero income tax—he says nurses making just $75K should save $12K a yearBy Preston ForeMay 21, 20262 days agoSuccessDespite a $500 million net worth, Shaq just finished his fourth degree. He warns graduates: 'Your character will take you further than your resume'By Preston ForeMay 20, 20263 days agoWorkplace CultureBolt CEO says he let go of his entire HR team for creating problems that didn’t exist: ‘Those problems disappeared when I let them go’ By Preston ForeMay 19, 20263 days agoSuccessIndeed chief economist says we’re entering an era of ‘great mismatch’ thanks to a generational imbalance of workersBy Emma BurleighMay 22, 202613 hours agoWorkplace CulturePay transparency is exposing a bigger problem: Most companies can't explain why they pay what they payBy Sydney LakeMay 20, 20262 days agoAIMicrosoft reports are exposing AI's real cost problem: Using the tech is more expensive than paying human employeesBy Jake AngeloMay 22, 202612 hours ago© 2026 Fortune Media IP Limited. All Rights Reserved. Use of this site constitutes acceptance of our Terms of Use and Privacy Policy | CA Notice at Collection and Privacy Notice | Do Not Sell/Share My Personal Information
FORTUNE is a trademark of Fortune Media IP Limited, registered in the U.S. and other countries. FORTUNE may receive compensation for some links to products and services on this website. Offers may be subject to change without notice.

Microsoft reports indicate that the use of artificial intelligence is more expensive than paying human employees, highlighting a significant cost problem associated with adopting this technology. This pressure is causing major technology firms to reassess their internal strategies regarding AI deployment. For instance, Microsoft has reportedly begun canceling most of its direct Claude Code licenses, shifting engineers toward using GitHub Copilot CLI, a move prompted by the widespread popularity and scale of AI experimentation. This action occurs while Microsoft maintains major investments, such as its Foundry deal involving Anthropic and Azure compute capacity. Similarly, other major corporations are implementing internal measures to manage AI spending; Uber's CTO reported exhausting its 2026 AI coding tools budget rapidly, and Meta utilized a leaderboard named "Claudeonomics" to track employee AI usage. Amazon is also pushing employees toward maximizing the use of AI tokens through a metric called "toxenmaxx."

These developments suggest that the economic feasibility of relying on AI for labor augmentation is more complex than initial optimistic forecasts suggested, echoing comments made by Bryan Catanzaro, vice president of applied deep learning at Nvidia, who stated that the cost of compute often exceeds the cost of employees. This difficulty in justifying the technology underscores a tension between the promise of an AI revolution and the actual economic costs of its adoption.

An emerging paradox exists concerning token-based pricing: although the cost of individual AI tokens is expected to fall sharply, aggregate costs can rise due to increased consumption and greater efficiency demands. Goldman Sachs previously forecasted that agentic AI could lead to a twenty-four fold increase in token consumption by 2030, potentially reaching a staggering twelve hundred quadrillion tokens per month. Although research from Gartner predicts that inference costs for large language models will decrease significantly by 2030, the report cautions that this deflation of commodity tokens does not equate to cheaper enterprise AI. Gartner warns that agentic models require substantially more tokens per task, and increased consumption could negate unit cost reductions, leading inference costs to potentially rise instead. This suggests that Chief Product Officers must distinguish between the deflation of base tokens and the democratization of advanced reasoning. This cost reality complicates broader expectations for deploying sophisticated AI agents, particularly as industry leaders anticipate a future where hundreds of AI agents operate alongside employees.