Uber Technologies has introduced new spending controls on employee use of AI coding tools after the company reportedly exhausted its annual AI budget within just four months, Bloomberg reports.
Under the new policy, employees are limited to approximately $1,500 per month for each AI coding platform, including tools such as Anthropic’s Claude Code and Cursor. Usage is monitored through an internal dashboard, although employees can request approval to exceed the limits in certain cases.
The move follows Uber’s aggressive push to encourage AI adoption across its workforce. Earlier this year, the company reportedly urged employees to maximize AI usage and even tracked participation through internal leaderboards.
However, the rapid increase in usage led to unexpectedly high costs. Uber’s leadership disclosed in April that the company had already consumed its full annual AI budget only a few months into the year.
The spending controls also highlight a broader challenge facing enterprises as AI adoption accelerates. While companies continue investing heavily in AI-powered development tools, automation platforms, and generative AI services, many are still struggling to quantify the direct business impact and return on investment.
Uber executives have acknowledged the uncertainty around AI productivity gains. Chief Operating Officer Andrew Macdonald recently noted that it remains difficult to directly connect AI usage to measurable product innovation or customer-facing improvements.
The development reflects a growing shift among enterprises from experimentation toward cost management and efficiency. In a similar move, Microsoft has reportedly begun canceling many internal Claude Code licenses and directing employees toward GitHub Copilot CLI, highlighting how rising AI coding tool expenses are prompting even major technology companies to reassess usage and spending strategies while continuing to invest heavily in AI infrastructure and automation capabilities.
