Uber named AI in its support cuts, so your team is already asking about theirs

Last Wednesday, Uber cut 10% of the jobs in its customer service operation and said AI was part of the reason. It was the first time the company has tied layoffs to AI, and anyone on your team who reads the news saw it.

Uber’s headcount is Uber’s business. What matters for you is what your own people are now wondering about their jobs, and whether they hear it from you first.

What Uber did and what it said

The cuts landed in Community Operations, the group that handles support across Uber’s marketplace. Uber doesn’t publish the size of that team, so there is no absolute number, only the percentage. Remote employees on the team were also told to move to a hub office, in line with Uber’s return-to-office rules, according to Bloomberg’s reporting.

An Uber spokesperson said the changes were meant to simplify operations, strengthen in-person collaboration and “continue to embrace AI.” Megha Yethadka, who runs global community operations, sent her division a memo that day. Her argument was that the team had grown too complex and siloed to get real value from AI, and that the structure had to change first. One line from it has been quoted widely: “We cannot scale frontier technology on top of fragmented processes.”

This came a month after Uber cut 23% of its people division, which was under one percent of its 34,000 employees, and two months after it said it would slow hiring because of its own AI use.

CMSWire went through the LinkedIn posts of six people who were let go. Between them they had more than 45 years at Uber, in roles like escalations, training and team leadership. These were not the newest hires on the scripted queue.

TimelineUber's last three months

MaySays AI use will slow hiring
JuneCuts 23% of its people division
Jul 22Cuts 10% of support, cites AI

Your team reads the same headlines you do

Uber is one story. The pattern behind it is what your employees feel. Challenger, Gray & Christmas, which tracks announced US layoffs every month, said on July 1 that employers attributed 14,029 June job cuts to AI. That was 31% of the month’s total and the fourth month in a row AI led the list of stated reasons.

By the numbers

31%

of announced US job cuts in June were attributed to AI

Source: challengergray.com

So when you tell your company you are rolling out an AI agent in support or finance, nobody hears “new software.” They hear the Uber headline. Some of them will start updating their resume that night. The best ones will do it first, because they have the most options.

There is a second reason to care. In my view, the cuts most likely to be reversed are the ones made in a hurry. CMSWire’s roundup points to Klarna and Commonwealth Bank of Australia, which both walked back AI-linked support cuts after service got worse. Gartner expects that by 2027 half of the companies that cut support staff because of AI will hire for similar roles again. If you lose your most experienced people to fear before you have even decided anything, you pay that cost without getting anything for it.

What people are actually asking

The questions are simple. People rarely ask them out loud, so it helps to know them in advance:

Is my job going away?
If it changes, who decides how?
Will I be the one training the thing that replaces me?
What happens to the person next to me?

Silence answers all four badly. If you say nothing, people assume the worst and assume you know more than you do. Usually you don’t know yet. Say that.

How to have the conversation

I’ve spent twenty years building companies, and I would rather give a team hard news than leave it guessing. This is how I would run the conversation.

Say what you are doing and what you are not. “We are putting an agent on order-status emails in support. We are not cutting the support team this year.” Only say the second sentence if it is true. If you can’t promise it, don’t. A broken promise about jobs costs you more trust than an honest “I don’t know yet.”

Talk about tasks, not roles. Uber’s memo talked about structure and process. That is the right level for leadership. For the people doing the work, get specific. Name the task the agent will take, how many hours a week it eats today, and what the person who did it will do with those hours. If you can’t answer the last part, you are not ready to announce.

Put the experienced people in charge of the agent. The people Uber let go included trainers and escalation specialists. Those are exactly the people you want writing the agent’s instructions and checking its output. They know the edge cases. Making them the reviewers turns the most anxious group into the owners.

Show them the agent before it goes live. Sit the team down and run it on twenty real tickets from last week. Let them see where it gets things right and where it gets them wrong. Fear grows when something is abstract. An agent that confidently tells a customer the wrong refund window is a lot less scary than the one people imagine, and the team will start pointing out fixes on the spot.

Don’t borrow Uber’s language. Words like “simplify,” “streamline” and “embrace AI” read as code for layoffs now, because that is how they have been used all year. Say the plain version. “Starting in September, the agent drafts replies to shipping questions and Maria approves them before they go out” is a sentence nobody has to decode.

Give a date for the next update. “We’ll talk again on the 15th, with what we learned from the first two weeks.” A date tells people the plan is real and that they will not hear about changes from a news story or a calendar invite from HR.

Ask the hard question yourself. If nobody asks “is this about cutting jobs,” ask it yourself and answer it. People trust a leader who says the quiet question out loud.

Before and after

What people hear in silence

AI is coming for my job
Leadership has already decided
I should start looking

What a clear plan tells them

The agent takes one named task
My role and hours are clear
A date for the next update

The same rollout lands very differently depending on what you say before it starts.

What to do this week

Write down every AI project you have running or planned, and the specific task each one touches. One line per project.
For each one, write what happens to the hours it frees up. Be honest. If the answer is “fewer hires next year,” say that to yourself first.
Hold a short meeting with the team closest to the first project. Tell them the task, the timeline and who will review the agent’s work.
Name one experienced person on that team as the reviewer, and give them time in their week to do it.
Put the next update on the calendar before you leave the room.

Uber may turn out to have made the right call for Uber. Your people will still read about it. Get to them before the headline does.