Uber Caps Employee AI Token Spending at $1,500 Per Month After Budget Exhaustion
The cap reveals that even a company deeply integrating AI into daily operations must impose hard limits once the costs outpace the budget, suggesting the productivity gains have not yet made unlimited spending sustainable.
Key Facts
- Uber set a $1,500 monthly cap per employee on AI coding tool token spending after exhausting its annual AI budget by early 2026.
- The cap applies separately to each AI tool, with a dashboard for employees to track usage and a process to request overage permission.
- CEO Dara Khosrowshahi said in February 2026 that software engineers use AI for about 90% of their work and about 30% of employees are AI power users.
- CTO Praveen Neppalli Naga said in April 2026 that about 11% of real-time backend system code updates are created by AI agents built with Claude Code, up from a few percent three months earlier.
Reporting from 1 source: GIGAZINE.
Uber Technologies has set a $1,500 monthly cap per employee on AI coding tool token spending after the company exhausted its annual AI budget early in 2026. The cap applies per tool, with a dashboard for tracking usage and a process to request overage permission. CEO Dara Khosrowshahi previously said engineers use AI for 90% of their work.
Uber Technologies has imposed a $1,500 monthly cap per employee on token spending for AI coding tools after the company burned through its annual AI budget by early 2026, a spokesperson told Bloomberg. The cap applies independently to each tool, and employees receive a dashboard to monitor usage. A process exists to request permission to exceed the normal limit. The move follows CEO Dara Khosrowshahi's February 2026 revelation that software engineers use AI for about 90% of their work and that about 30% of employees are AI power users. CTO Praveen Neppalli Naga said in April that about 11% of real-time backend system code updates are now created by AI agents built with Claude Code, up from just a few percent three months earlier.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.