Uber's operations chief is raising a red flag about a trend that has swept through corporate America: spending heavily on artificial intelligence without a clear way to measure what it is actually delivering. 

Business Insider reported that in a new interview, Andrew Macdonald said Uber is finding it "harder to justify" the cost of AI tools when there is no clear link between increased usage and more useful features for customers. 

Speaking in a Rapid Response interview released Saturday, Macdonald said internal conversations at Uber intensified after CTO Praveen Neppalli Naga said in an April interview with The Information that the company had exhausted its 2026 Claude Code budget. 

Macdonald described that moment as "head-exploding," saying it prompted broader discussions about AI token usage and what the company might be sacrificing to cover the cost, including staffing trade-offs. 

After talking with senior engineering leaders, he said he came away unconvinced that more AI spending was translating into a comparable jump in product improvements. In his view, there is not yet a clear line between higher token use and more useful customer features. 

His comments come after Uber CEO Dara Khosrowshahi said earlier this month on an earnings call that the company was cutting back on hiring, in part to balance its AI spending. 

Uber's comments reflect a more cautious view at a time when many companies are pushing workers to use AI as much as possible. In some corners of the tech industry, "tokenmaxxing" — maximizing AI use — has become a badge of productivity. 

AI systems also depend on massive data centers that draw electricity from the grid and often require large amounts of water for cooling. 

The technology is promising in some applications, but can also be expensive, resource-intensive, and vulnerable to misuse, security risks, or unintended consequences such as rising energy demand that can put pressure on utility bills. 

Uber's skepticism suggests more companies may start asking whether AI is actually solving meaningful problems — or simply adding another layer of cost. 

Some companies are already reassessing how aggressively they push AI internally. Uber appears to be scrutinizing whether usage metrics are the right way to judge value, especially if they do not translate into better products for riders, drivers, or delivery users. 

Duolingo has also shown signs of a pullback. After employees questioned whether they were being pressured to use AI just for its own sake, the company reversed its plan to factor AI use into performance reviews. 

"The link is not there yet, right?" Macdonald said, according to Business Insider. "It's very hard to draw a line between one of those stats and, 'Okay, now we're actually producing 25% more useful consumer features.'" And as Duolingo CEO Luis von Ahn put it, "in some cases [AI] did not fit."

A post of the article by Ed Zitron (@EdZitron) on the social platform X drew plenty of user feedback.

"I mean yeah obviously replacing Uber Eats menu item photos with AI slop isn't going to be useful," one commented.

"First time a company said the quiet part out loud," another added. "The AI spend isn't even tying back to anything useful, just vibes and invoices at this point."

Get TCD's free newsletters for easy tips, smart advice, and a chance to earn $5,000 toward home upgrades. To see more stories like this one, change your Google preferences here.