Uber imposes $1,500 monthly cap on AI tools after budget exhaustion

Uber has introduced strict internal spending limits on artificial intelligence tools, capping expenditure at $1,500 per employee per month for specific agentic coding platforms. The new policy, reported by Bloomberg, applies to tools such as Anthropic’s Claude Code and Cursor, marking a significant shift in the company’s approach to technology adoption.

The decision follows revelations that the ride-sharing giant exhausted its entire annual AI budget within just four months. According to The Information, this rapid depletion occurred during a period when staff were encouraged to utilise AI “as much as possible” and were ranked on internal leaderboards to drive engagement.

Usage of these tools is now monitored via an internal dashboard accessible to employees. While the $1,500 monthly limit per employee per tool is the standard, the company notes that exceptions can be granted with permission in certain cases, allowing for flexibility where specific operational needs arise.

Uber’s chief executive, Andrew Macdonald, has publicly questioned the return on investment of these tools. Speaking during a recent podcast appearance, Macdonald noted that it is “very hard to draw a line” between AI usage and the development of new consumer features, casting doubt on the direct productivity impact of the technology.

The move reflects a broader challenge facing the tech industry, where enterprises are investing heavily in AI with uncertain financial returns. As companies pour capital into these systems, the return on investment remains largely theoretical, leading to growing restlessness among organisations waiting for tangible results to justify the spend.

Uber’s cutback highlights the tension between aggressive early adoption and fiscal discipline. While the company previously promoted widespread AI integration, the current restrictive measures suggest a recalibration of expectations regarding the immediate value generated by agentic coding platforms.