Uber poured its entire 2026 AI budget into just four months, using Claude Code and Cursor. By April, the bill had consumed the full yearly allocation, forcing leadership to reassess. Engineers reported monthly API costs between $500 and $2,000 per person, while 95 % of the team relied on the tools. The surge highlighted the tools’ undeniable value for engineering productivity today.

Claude Code launched to Uber engineers in December 2025, doubling usage by February as developers discovered its multi‑step capabilities. By contrast, Cursor’s adoption plateaued. The CTO noted the company is now “back to the drawing board” on AI budgeting, questioning whether the $3.4 billion R&D spend can sustain such high tool costs. The rapid burn signals a shift in cost models today.

The incident underlines how AI tools can outpace budget forecasts, turning productivity gains into budgetary crises. With 95 % of engineers now using Claude Code monthly, limiting access feels counterproductive. Other firms likely face similar challenges as developers embrace high‑value coding assistants. Uber’s experience forces a reevaluation of how companies balance cost with engineering velocity for future development efficiency and innovation today.