A growing roster of public companies is blaming artificial intelligence for staff cuts. The pattern is hard to miss: name a big tech firm, and there is a fair chance it has trimmed headcount in 2025 while pointing at AI. According to reporting from businessinsider.com, Uber, Snap, Block, GitLab, Atlassian, Cisco, and others have all cited the technology, directly or indirectly, when announcing reductions.
For project managers, team leads, and aspiring agile practitioners, the question is not philosophical. It is operational. If your company follows, do you have a plan beyond the press release?
What happened:
- GitLab said in early June it would cut 350 employees, framing the move as preparation for what it called the "agentic era."
- Uber laid off roughly 10% of its customer service staff in July, citing AI.
- Block's Jack Dorsey said in February he was cutting the company's headcount from over 10,000 to under 6,000, citing new AI-driven ways of working.
- Atlassian announced 1,600 cuts in March, about 10% of staff, with CEO Mike Cannon-Brookes naming AI and enterprise growth as drivers.
- Cisco said on May 13 it would cut fewer than 4,000 jobs, with CEO Chuck Robbins linking the reorganization to AI strategy.
- A March Challenger, Gray, and Christmas report found AI cited in 8% of job-cut plans so far this year. A 2025 Robert Half survey said 29% of 2,000 hiring managers had reopened roles after trying AI.
The honesty problem: AI or excuse
Be blunt about the first issue. There is likely some amount of AI washing in the mix. OpenAI's Sam Altman has already said publicly that some companies blame AI for cuts that would have happened anyway. There is also real pressure to show a generative-AI story to Wall Street, and a layoff is an easy way to attach a productivity narrative to cost discipline.
The numbers support skepticism. An MIT study released last year said 95% of corporate AI investments have generated zero return so far. The story is not "AI is replacing everyone." It is "AI is a useful narrative during a restructure."
How we got here: the post-2022 correction
The macro story is well known. After a hiring binge during the 2020-2022 growth period, tech operating models hit slower revenue and higher rates. Middle management layers, duplicated support functions, and pre-2023 customer service build-outs all looked expensive. AI arrived at the exact moment companies wanted a reason to act. It is the rare technology whose hype directly justifies headcount decisions, which is why it shows up in almost every recent memo.
The wrinkle in 2025 is that AI capability actually moved. It is no longer a slide. That changes which roles get cut first, and which get rehired.
Our analysis
Here is the read for anyone running delivery. AI layoffs are not random. They tend to cluster in three areas: customer support, mid-level engineering tasks that look repeatable, and back-office operations. They tend to leave alone the people who can ship AI products, sell them, or lead the teams that do.
That means the trade-off is sharper than CEOs admit in public. You lose institutional knowledge and on-call muscle. You keep the people who can prompt, prototype, and integrate. Over a two-year horizon, the winners are the firms that keep enough of the old guard to run reliable operations while the new shape stabilizes. The losers are the firms that cut too deep and discover, as Robert Half's survey hints, that reopened roles cost more than retained ones.
The second-order risk is culture. Cannon-Brookes said he expects more engineers in five years, not fewer. That is the bet every CEO making these cuts is making. If the bet is wrong, the survivors will remember who made it.
A worked example for your team
Say you are a delivery lead running two squads and you hear your VP is considering an AI-driven restructure. Do not wait for the memo. Pull your last four sprints and tag tasks by automation potential. Identify which ceremonies are now redundant, which review steps AI can pre-fill, and which two or three people on your team hold knowledge nobody else does. Walk into the conversation with a written answer to "what should this team look like in 12 months?" Teams that show up with a concrete plan tend to be the ones kept intact, even when the org around them shrinks.
What to watch next
Bottom line: AI is real, but the layoff wave is more about post-2022 right-sizing with a convenient label. Watch the rehiring data. If firms that cut in Q1 and Q2 of 2025 start posting the same roles in late 2025 or early 2026, the AI narrative collapses on itself. The leaders who come out ahead will be the ones who treated this as a redesign, not a reduction.
Key takeaways
- Treat AI layoff memos as signals of cost discipline plus AI branding, not pure automation math.
- Map your team's work by automation potential before any restructure talk reaches you.
- Keep the people who hold undocumented institutional knowledge; they are the hardest to rebuild.
- Track rehiring announcements. They are the clearest tell on whether the AI bet paid off.
- Bring a written 12-month team plan into leadership conversations; reactive teams get downsized.
Frequently asked questions
Is AI really causing these layoffs, or is it just a convenient story?
It is both. Some roles are genuinely being automated, especially in customer support and repeatable back-office work. Other cuts were likely planned for cost reasons and rebranded with an AI label for board and market optics. The honest answer is that AI gives leaders a clean narrative during a restructure that was already overdue.
Which roles are getting cut first in the 2025 AI layoffs?
Customer support, mid-level repeatable engineering, and back-office operations. Companies are protecting revenue-generating roles tied to AI products, senior engineers who can build agentic systems, and customer-facing sales staff. Cloudflare's CEO, for example, said very few engineers or customer-facing salespeople were affected by its cuts.
What should a project manager do if their company hints at an AI restructure?
Do a sprint-level audit now. Tag tasks by what AI could automate, identify ceremonies and reviews that are redundant, and list the two or three people on your team who hold undocumented institutional knowledge. Walk into leadership conversations with a concrete 12-month team plan rather than a defensive reaction.
Will companies that cut staff over AI have to hire them back?
Likely some of them. A 2025 Robert Half survey of 2,000 hiring managers found 29% had reopened positions they had previously eliminated after trying AI. The pattern shows up most often when companies cut too deep into operational roles that AI turned out not to handle reliably.
What should leaders watch over the next six months?
Three things. First, rehiring announcements from the companies that cut in early 2025. Second, any CEO reversal on AI headcount claims. Third, whether AI-product revenue actually grows in the firms that bet the hardest on the narrative. If revenue does not follow, the layoffs look even more like AI washing in hindsight.