Uber has introduced a monthly spending limit on employee use of AI coding tools, becoming one of the latest technology companies to tighten oversight of artificial intelligence costs as adoption accelerates across the business.
According to a report by Bloomberg, the company has set a $1,500 monthly spending cap per employee for each AI coding platform, including tools such as Claude Code and Cursor. The move follows reports that Uber consumed its entire annual AI budget during the first four months of 2026.
The restrictions mark a notable shift for a company that had previously encouraged employees to integrate AI extensively into their daily workflows.
New controls target rapidly growing AI usage
The spending limits apply to so-called agentic coding tools, AI systems capable of writing, reviewing and modifying software code with limited human intervention.
According to Bloomberg, Uber said the new policy is designed to encourage responsible experimentation while ensuring costs remain manageable as AI usage expands across the organisation.
Key elements of the policy include:
- A $1,500 monthly budget per employee for each AI coding tool
- Separate spending limits for each platform used
- Approval processes for employees requiring additional access
- Internal dashboards allowing workers to monitor AI usage and spending
- Controls focused on agentic AI coding applications
Employees who use multiple AI platforms receive separate budgets for each service rather than a single combined allowance.
Budget concerns emerge after aggressive AI push
The spending restrictions come only months after Uber actively promoted AI adoption among employees.
According to previous reports cited by Bloomberg, staff had been encouraged to use AI tools extensively, with internal leaderboards ranking employees based on their usage levels.
In April, Chief Technology Officer Praveen Neppalli Naga disclosed that Uber had already exhausted its annual AI budget during the first four months of the year.
The development illustrates a growing challenge facing large enterprises. While AI tools often deliver productivity gains, usage-based pricing models can cause costs to rise rapidly as adoption spreads across teams and departments.
AI becomes embedded across Uber's operations
Uber's leadership has repeatedly emphasised the role AI now plays across the company.
Chief Executive Officer Dara Khosrowshahi recently said that approximately 10% of Uber's code is generated and submitted by AI agents.
The technology is no longer confined to engineering teams. According to company executives, legal, marketing and other business functions have increasingly adopted AI-powered tools to automate tasks and improve productivity.
The expansion reflects a broader trend among technology companies that are integrating AI into everyday workflows rather than limiting it to specialist teams.
Measuring business value remains difficult
Despite strong adoption levels, Uber executives have acknowledged that evaluating the long-term impact of AI remains challenging.
Speaking on the Rapid Response podcast, Chief Operating Officer Andrew Macdonald said it remains difficult to directly connect increased AI usage with the delivery of better customer-facing features.
While internal productivity indicators have improved, Macdonald noted that quantifying the broader business value generated by AI remains a work in progress.
The comments highlight a wider debate across the technology industry, where many organisations are balancing rising AI investment against uncertain measures of return on investment.
Uber joins wider industry effort to manage AI costs
Uber's move comes amid broader efforts by technology companies to control spending on AI infrastructure and software tools.
According to reports, Microsoft has also begun restricting internal access to certain AI coding platforms. The company has reportedly limited employee access to Anthropic's Claude Code and directed engineers in its Experiences + Devices division to move towards GitHub Copilot CLI.
The reported transition affects teams responsible for products including Windows, Microsoft 365, Outlook, Teams and Surface devices.
Industry reports suggest the decision was influenced in part by the operational costs associated with agentic AI systems, which can consume large volumes of computing resources through complex, multi-step workflows.
As enterprises move from AI experimentation to large-scale deployment, spending discipline is emerging as a new priority. Uber's latest restrictions underline a reality increasingly facing technology leaders: while AI may boost productivity, scaling its use across thousands of employees can generate significant and ongoing costs.