Uber Flags Rising Difficulty in Defending AI Token Outlays

*Uber’s chief operating officer Andrew Macdonald said the company is finding it harder to justify continued spending on AI tokens.*

The statement

Uber’s COO Andrew Macdonald told Business Insider that token-related AI costs are becoming tougher to defend internally. The remark was highlighted on the Hacker News front page on 25 May 2026, where it drew 104 points and 114 comments.

Limited details released

No specific figures or timelines accompanied the comment. The company has not disclosed current monthly token consumption or the models involved. Prior public updates from Uber on its AI usage have been limited to high-level product announcements rather than infrastructure accounting.

Industry pattern

Other large technology firms have begun publishing more granular breakdowns of inference spend. Uber’s position suggests it may follow that pattern or apply stricter internal thresholds for new AI projects. The COO’s phrasing leaves open whether existing workloads will face cuts or simply slower expansion.

Why it matters

For teams that rely on Uber’s mapping, logistics, or marketplace platforms, the comment signals a possible slowdown in new AI-driven features whose value is hard to quantify against token bills. Companies watching similar spend lines will treat the remark as an early indicator that return-on-inference calculations are tightening across the sector. The next earnings call or infrastructure update will show whether Uber translates the statement into measurable restraint.

---

Sources:

• Business Insider

{
  "excerpt": "Uber’s COO says the company is finding it harder to justify spending on AI tokens.",
  "suggestedSection": "business",
  "suggestedTags": ["uber", "ai-costs"],
  "imagePrompt": "Abstract geometric tokens stack unevenly on a sparse industrial floor while thin lines of light trace their connections. Some stacks cast long shadows across the surface. muted color palette, cinematic lighting, 16:9"
}