Uber cuts 10% of support staff for AI — but its own memo reveals the real bottleneck

The facts first. Uber cut roughly 10% of jobs in its community operations team — the group that handles customer service — according to a Bloomberg report cited by NDTV Profit. An Uber spokesperson framed the move as an effort to "simplify operations, strengthen in-person collaboration, and continue to embrace AI." Remote workers on the team were also told to relocate to a hub office under a return-to-office mandate.

The number itself is modest in context. Uber employs about 34,000 people globally, and the company hasn't disclosed the size of its customer-service unit. This follows a June cut of 23% of its "people" division — which the company said represented less than 1% of total staff — and marks the second round of reductions in under two months. It is, however, the first time Uber has explicitly tied job cuts to AI efficiency, joining Block and Oracle in citing the technology as a reason for shrinking teams.

The most revealing detail isn't the layoff figure — it's a memo from Megha Yethatika, Uber's VP of global community operations, quoted in the report. She wrote that the organization "has become too complex and siloed," and that while the team "has made some strides" with AI, "we cannot scale frontier technology on top of fragmented processes." That sentence deserves to be read twice.

Here's the context worth remembering. Customer service is exactly the kind of high-volume, rule-based, back-office function that our sector analysis has repeatedly flagged as the most exposed to automation. Basic support tickets are cheap, repetitive, and pattern-heavy — the natural first target. This is consistent with what we've seen elsewhere: the routine goes first, the judgment-heavy and relational work resists longer.

But the transparency in Yethatika's memo cuts against the easy "AI replaced the workers" narrative. She is admitting that the technology didn't just walk in and do the job. It required reorganizing messy, siloed workflows first. That's the recurring lesson of this cycle — the competitive edge is shifting from having a model to integrating it into real operations. A capable model dropped onto broken processes delivers little. And we'd add a note of caution: nearly half the companies that cut staff citing AI have been quietly rehiring, sometimes at higher cost, when the automation proved less turnkey than the press release implied.

Our reading: this is a genuine transition, and it's uneven. For the people affected, the disruption is real and shouldn't be softened with slogans. But the story here is less "AI is eliminating jobs" and more "companies are using AI as the occasion to flatten bloated org charts they'd wanted to cut anyway." The honest version — the one in Uber's own memo — is that automation is a forcing function for organizational cleanup, not a magic replacement. The long-term horizon still points toward machines absorbing the drudgery so humans do higher-value work. The near-term reality is messier, and the companies pretending otherwise are the ones most likely to be rehiring next year.