On Monday April 27, three independent layoff trackers crossed the same line within hours of each other. None of them held a press conference. The number is 100,000.
SkillSyncer’s tracker showed 155 layoff events impacting 100,443 workers in 2026, with the most recent update timestamped that morning. TrueUp.io had 254 tech-company layoffs impacting 104,093 workers through April 24. Peerlist’s tracker listed 144 companies and roughly 102,000 employees for the year. Three different methodologies, three different denominators, same ballpark: the U.S. tech industry has cut roughly 100,000 jobs in the first 117 days of 2026.
That is 858 layoffs per day. Every single calendar day, including Sundays, including the four federal holidays that fell in the window. If you started reading this article at the top, by the time you finished, six tech workers somewhere were no longer employed.
What “AI-attributed” actually means now
The number that moved the most this year is not the headcount cut. It is the share of those cuts that gets blamed on AI in the announcement.
Tom’s Hardware reported that 47.9% of Q1 2026 tech layoffs — 37,638 of 78,557 cuts — were explicitly attributed to AI and workflow automation by the laying-off company. A year earlier, that share was below 8%. Three months earlier, it was around 20%. The graph is not a curve; it is a step function.
Two things are doing the work here:
- AI actually does some of the labor now. Customer-support volumes routed to LLM agents, code review handed to Copilot or Claude or Cursor, legal-document review handed to Harvey or Hebbia, accounting reconciliation handed to internal agents — all of this is real, all of it has shipped, and all of it has measurable per-ticket / per-PR / per-document cost reductions on the books of the companies citing the number. Citi told its Q1 call that 80% of staff have onboarded AI tools and the bank logged 42 million AI interactions last quarter.
- AI also “plays better with stakeholders.” A December 2025 survey of 1,000 hiring managers found 59% admit they emphasize AI in layoff press releases because it sounds better than admitting demand is soft, gross margins are compressing, or the CFO needs Q2 EPS to clear consensus. Both motivations are present in the same announcements.
The honest read of 47.9% is: AI is doing some of the work that used to be done by some of the people who got laid off, and also it is a more flattering quarterly-call narrative than the alternative explanations. Both can be true. They are.
What landed on the trackers in the last 17 days
The clustering matters. From Fast Company’s April tracker and CNBC’s wrap-up:
- Oracle — confirmed 10,000 cuts on April 1, with internal expectations the wave expands toward 20,000–30,000.
- Meta — 8,000 cuts (10% of headcount) announced April 17 with cuts beginning May 20.
- Microsoft — first-ever voluntary buyout offered to ~7% of US headcount (roughly 8,750 people) on April 24. The first buyout in the company’s 51-year history. AI and Copilot teams are explicitly exempt.
- Snap — 1,000 cuts (16% of staff) on April 22.
- Nike — 1,400, mostly in technology, on April 25.
- Disney — 1,000 across studios, ESPN, and product/technology on April 21–22.
- Wells Fargo, Bank of America, JPMorgan, Goldman, Morgan Stanley, Citi — collectively shed ~15,000 in Q1, posted record $47B in profits.
Three weeks. Roughly 35,000 confirmed cuts at named companies. Not counting the dozens of <500-person reductions that don’t make the trackers’ headline list.
The story underneath: buyouts, not layoffs
The most consequential structural shift this month is not the count. It is Microsoft choosing buyouts over involuntary layoffs, and Fortune’s April 26 explainer mapping out the reasons.
Voluntary buyout programs have three properties an involuntary RIF does not:
- No WARN notice, no reputational hit. The headline reads “Microsoft offered employees a voluntary exit,” not “Microsoft fired 8,750 people.” The first reads as benevolent; the second tanks the engineering-employer brand.
- Self-selection on who leaves. The people most likely to take a buyout are senior, well-compensated, and within a few years of retirement — exactly the people whose chairs the company can refuse to refill. The people who decline are the ones whose performance reviews say the company should keep them anyway.
- Operating-margin disclosure flexibility. Severance from a voluntary program is a one-time charge that gets normalized out of “adjusted” earnings; the steady-state savings show up in subsequent quarters as margin expansion. Wall Street loves this.
If Microsoft’s buyout works — and the early read from Wall Street was “yes, the stock liked it” — then expect Apple, Amazon, Alphabet, Salesforce, Adobe, and the next tier down to follow with their own first-ever buyout programs over the next two quarters. The “involuntary RIF” becomes the option you only use when the voluntary program under-fills.
Pichai’s 75% as the missing context
Three days before Microsoft’s buyout, Sundar Pichai told Cloud Next 2026 that 75% of new code at Google is AI-generated and engineer-approved, up from 25% in October 2024. The doubling time is roughly six months.
That number is the operating leverage underneath the layoff trackers. If 75% of one of the world’s most-watched engineering org’s new code is being written by an agent, the unit economics for “how big does the engineering team need to be to ship the same roadmap” change inside every CFO’s spreadsheet, not just Google’s. The trackers are the output. Pichai’s number is the input.
The bridge between the two is roughly nine months. The 25%-October-2024 figure preceded the Q4-2024 layoff wave by about a quarter. The 50%-fall-2025 figure preceded the Q1-2026 wave. The 75%-April-2026 figure points at Q4 2026 and Q1 2027. The trackers do not know this yet; the spreadsheets do.
What LostJobs is watching
- Whether the May Challenger report shows AI cited above 30% of total announced cuts. March came in at 25% — already the highest single-month share Challenger has ever recorded. April will publish in early May. If it crosses 30%, AI moves from “leading reason” to “majority reason” in a single quarter.
- Whether any other Big-Tech CEO matches Microsoft’s buyout structure by end of Q2. Buyouts vs. layoffs is now the operating choice. The first three Fortune-100 boards to adopt the Microsoft template ratify it as the default playbook for 2026–2027 reductions.
- Whether the BLS unemployment series catches up to the tracker series in the next two prints. The trackers measure announcements; BLS measures labor-force status. The gap between “announced” and “filed for unemployment” runs about 60–90 days. The Q2 BLS prints (May 2 and June 6) will tell us whether the announcement wave converted into measurable unemployment, and whether the Bloomberg editorial board’s “AI apocalypse delayed” thesis survives contact with the actual data.
The dry coda: the most-circulated screenshot on engineering Twitter on Monday was not the 100,000 number. It was a Layoffs.fyi tracker entry showing the headcount counter ticking past six figures, with the timestamp “2026-04-27 06:47 PT.” The reply with the most likes was: “Ten years ago this would have been a national news event. Today it cleared the line on a Monday morning before the Bay Area woke up.” Both halves of that sentence are the actual story.