Uber caps employee AI spending after blowing through budget in 4 months

Uber burned through its entire annual AI budget in four months. After initially encouraging employees to use AI tools as much as they wanted, the company is now implementing spending caps per team.

This is a perfect example of how quickly AI costs can spiral. When you tell thousands of employees to use Claude, ChatGPT, or internal AI tools freely, those API calls and compute costs add up faster than most finance teams anticipate.

The irony here is notable. Uber went from actively pushing AI adoption to suddenly pumping the brakes, which probably caused whiplash for teams that had built AI into their workflows.

For anyone managing AI tools at their company, this is your warning sign. Set usage policies and budget guardrails from day one, not after you've blown through your allocation. Track costs by team and use case so you can spot expensive patterns early.

The shift from unlimited to capped also signals something bigger. Companies are moving past the experimental phase where they throw money at AI to see what sticks. Now comes the harder part: figuring out which AI uses actually deliver ROI and which are just expensive novelties.

Uber hasn't shared specifics on the new caps or what drove the overspend. But if a company built on optimizing costs and margins can miscalculate AI spending this badly, it can happen to anyone.