Uber Says Its AI Costs Just Aren’t Worth It

Uber’s AI spending is under scrutiny after the company’s COO, Andrew Macdonald, said during a podcast appearance that the financial benefits of AI tools have not materialized. He noted that while AI usage is rising, Uber struggles to link the costs to measurable improvements in consumer features. Macdonald referenced Uber’s CTO, Praveen Neppalli Naga, who earlier admitted the company’s 5,000 engineers burned through its 2026 Anthropic Claude Code token budget by March. Microsoft also faced similar cost concerns, discontinuing Claude Code licenses in favor of its in-house Copilot tool, reportedly due to financial pressures. Employees suggested the move was driven by budget constraints rather than operational efficiency. Despite these challenges, Uber remains committed to AI, with CEO Dara Khosrowshahi stating in an earnings call that AI adoption is enhancing employee productivity across legal, marketing, and development teams. However, the company is slowing hiring to manage rising AI expenses. The shift reflects broader industry concerns about AI’s economic viability, as companies question whether the high costs align with tangible returns. Macdonald warned that without clear productivity gains, justifying AI spending becomes difficult, especially as cloud tool prices surge. Uber’s AI investments continue, but the company’s leadership is now evaluating whether the financial trade-off remains sustainable. The debate highlights growing skepticism about AI’s cost-effectiveness in tech-driven organizations.