Uber caps employee AI spending after blowing through budget in four months

Uber has imposed a $1,500 monthly spending cap per employee on agentic coding tools after exhausting its annual artificial intelligence budget in four months, according to a Bloomberg report [1]. The ride-hailing and delivery company, headquartered in San Francisco [5], now limits each staff member to $1,500 per month on tools such as Anthropic’s Claude Code and Cursor [1]. Employees can track their usage through an internal dashboard, and the cap can be exceeded with permission in certain cases [1]. The restriction follows a disclosure by Uber’s chief technology officer in April that the company had consumed its full-year AI allocation within the first third of the fiscal period [1]. Prior to the budget overrun, Uber had urged staff to use AI “as much as possible” and ranked internal adoption on competitive leader boards, The Information previously reported [1]. The aggressive promotion of these tools contributed to the rapid depletion of funds, prompting the new controls. Uber CEO Andrew Macdonald expressed skepticism about the productivity gains from AI during a recent podcast appearance. “It’s very hard to draw a line” between AI usage and the delivery of new consumer features, Macdonald said [1]. His remarks underscore a tension that has emerged across the technology sector as companies increase AI investment without clear evidence of proportional returns. The San Francisco Bay Area, where Uber is based, is among the most economically productive large urban areas in the world, with a GDP of $1.332 trillion in 2024, driven in part by its concentration of technology firms [5]. Uber shares the city with companies including Salesforce, Airbnb, OpenAI, and Anthropic, the maker of Claude Code [5]. Anthropic is one of the AI providers whose tools are now subject to Uber’s internal spending limits [1]. Uber’s belt-tightening arrives as enterprises across the industry confront questions about AI return on investment. While spending on generative AI and coding assistants has accelerated, measurable productivity improvements remain elusive for many organizations [1]. The move to cap per-employee tool expenditure signals that even well-capitalized technology firms are beginning to scrutinize the cost side of AI adoption more closely [1].