Companies spent the past two years racing into artificial intelligence because nobody wanted to be seen falling behind. Now the bills are piling up, finance teams are pushing back and some executives are struggling to prove the spending is actually improving the business.
Uber revealed earlier this year that it had already exhausted its entire annual AI budget within the first few months of 2026 as employees rapidly adopted coding tools and AI systems across the company. Executives later admitted it remained difficult to clearly measure whether the surge in spending was translating into noticeably better products or stronger consumer performance.
That tension is no longer isolated to one company. Across the technology sector and corporate America more broadly, businesses are beginning to confront the financial weight of an AI race many felt they had little choice but to join. Companies including Meta, Amazon and Disney have reportedly been tracking internal AI usage aggressively as management teams push staff to adopt the technology faster. In some workplaces, employees are even being ranked based on how heavily they use AI systems.
The result is creating some strange behavior inside offices. Some workers are reportedly running unnecessary AI tasks simply to demonstrate adoption, while companies continue pouring money into systems they are still struggling to properly evaluate.
Inside some companies, the mood has already changed. Investors still want visible AI progress. But many management teams are now under pressure to show the spending is producing something tangible rather than simply driving up costs.
Most consumers experience AI through free chatbots or low-cost subscriptions. Large businesses face something very different. Their usage is increasingly metered at scale, with costs tied to the number of AI “tokens” consumed as systems generate code, analyse data and automate workflows. Once multiple AI agents begin operating continuously across departments, the spending can escalate quickly.
The spending itself is becoming harder to ignore. Meta recently cut thousands of jobs while ramping up AI investment, while Uber acknowledged that heavier AI spending is partly being offset by slower recruitment. Across white-collar industries, many workers no longer see AI as just another software tool. It is becoming tied to hiring decisions, budgeting priorities and growing questions about long-term job security.
Some companies are already becoming more cautious after the first wave of AI enthusiasm. Businesses are reassessing whether constant AI usage improves productivity enough to justify the cost, particularly as token prices rise and more advanced systems demand heavier computing power. That is where the conversation inside boardrooms is beginning to shift.
The focus is moving away from experimentation and toward cost control, efficiency targets and whether the promised productivity gains can truly support the scale of spending now building across the corporate world.
Few major businesses appear willing to slow their AI investment plans. But executives who once feared falling behind are beginning to ask a different question: how long can companies keep spending at this pace before the strain starts showing up somewhere else?
