Uber Caps Employee AI Spending: Budget Blown in Just Four Months

The escalating costs associated with artificial intelligence are prompting various companies to curtail their usage in an effort to manage expenditures. Among these organizations is Uber, which has recently implemented internal usage limits to mitigate its substantial AI-related spending.

According to a Bloomberg report, Uber has established a new policy imposing a monthly expenditure ceiling of $1,500 per employee for each agentic coding tool utilized, including platforms like Anthropic’s Claude Code or Cursor. Employees can monitor their consumption through an accessible internal dashboard, with the company noting that these caps may, in specific circumstances, be surpassed with prior authorization.

This development is perhaps unsurprising, given that Uber's Chief Technology Officer disclosed in April that the ride-sharing giant had exhausted its entire annual AI budget within just four months. This rapid depletion reportedly occurred subsequent to Uber's earlier encouragement for staff to maximize AI usage, even implementing competitive internal leaderboards to track adoption, as previously reported by The Information.

Adding to the discourse, Uber's CEO, Andrew Macdonald, recently voiced skepticism regarding AI's tangible impact on productivity. During a podcast appearance, he remarked that it is "very hard to draw a line" connecting AI utilization directly to the creation of new consumer features.

Uber's decision to scale back underscores a more pervasive challenge confronting the technology sector: as businesses funnel substantial investments into artificial intelligence, the precise return on investment remains elusive. Indeed, the ROI of AI has largely persisted as a theoretical concept that many hope will eventually materialize, though it is evident that some companies are growing increasingly impatient awaiting its realization.