Uber President Casts Doubt on AI Spending ROI Amid Budget Strain
· Jess Weatherbed
Uber blew through its entire 2026 AI budget just four months into the year. Now the company is publicly questioning whether the money is actually buying anything useful.
Uber blew through its entire 2026 AI budget just four months into the year. Now the company is publicly questioning whether the money is actually buying anything useful.
President and COO Andrew Macdonald told the podcast Rapid Response that Uber has yet to see a clear connection between rising token consumption for its internal AI tool Claude Code and any real improvements for riders or drivers. “That link is not there yet, right?” he said. He acknowledged that maybe more features are getting shipped, but admitted it’s “very hard to draw a line” between the stats and actual consumer results.
The honest admission stands out in an industry where companies rarely admit they might be overspending on AI. Uber’s problem isn’t unique. Many firms are pouring cash into AI tools without a clear metric for return on investment. Token consumption, the measure of how much AI processing Uber’s engineers are using, keeps climbing, but Macdonald is essentially saying the company doesn’t know what it’s getting in exchange.
The takeaway here is less about Uber specifically and more about a pattern. Massive AI spending is happening in good faith, often without proof that it’s producing better products. If a company as data driven as Uber can’t draw that line, maybe the rest of the industry shouldn’t assume the line exists either.