AI Job Cuts in 2026 Already Beat All of Last Year
Challenger, Gray & Christmas data shows 87,714 US jobs have cited AI as a cut reason through May 2026 - already past all of 2025's total - while companies post record profits.
7 min read
Employers in the United States cited artificial intelligence in connection with 87,714 job cuts through May 2026 - a total that has already surpassed every AI-attributed cut recorded in all of 2025, when the full-year count came to 54,836.
TL;DR
- 87,714 US job cuts cited AI through May 2026 - already past the 54,836 recorded across all of 2025
- AI became the top-cited reason for layoffs for the third consecutive month in May (Challenger, Gray & Christmas)
- 38,579 cuts in May alone named AI - the highest monthly total since tracking began in 2023
- Nine companies each cut 10,000+ jobs while simultaneously reporting record revenues or heavy AI investment
- Self-reported attribution means companies decide how to frame cuts - skepticism is warranted
The data comes from Challenger, Gray & Christmas, the US outplacement firm that has tracked employer-stated reasons for job cuts since the 1990s. Their May 2026 report, released this month, confirmed that AI led all stated reasons for the third month running, accounting for roughly 40% of May's total planned cuts. The figure for the year is already 60% higher than the equivalent full-year 2025 number with seven months still to go.
The Numbers
The scale is easier to grasp company by company. Nine employers have each announced more than 10,000 positions tied to AI or automation reasons: Amazon, Microsoft, Accenture, Citigroup, Dell, HSBC, Intel, TCS, and UPS. Several of them reported earnings highs in the same quarters they announced the cuts.
| Company | Cuts | Executive framing |
|---|---|---|
| Amazon | 30,000 | "Advances in AI allow the company to operate more efficiently with fewer people" |
| Microsoft | 23,000 | $100B+ annual AI investment; workforce realignment |
| Oracle | 21,000 | $3.7B quarterly net income, up 27% YoY; savings redirected to AI data centers |
| Standard Chartered | 7,800 | Replacing "lower-value human capital" with machines by 2030 |
| Meta | 8,000 | 7,000 simultaneously reassigned to AI roles |
| Salesforce | 5,000 | "AI agents now handle about 50% of customer interactions" |
| Cisco | 4,000 | Record profits; realigning resources to AI and security |
| Block | 4,000 | Workforce cut from 10,000+ to under 6,000 |
| Cloudflare | 1,100 | "AI had made these roles obsolete" |
| GitLab | 350 | Exited 22 countries; restructured for the "agentic AI era" |
The month-by-month acceleration is visible in Challenger's own numbers: AI cuts in 2025 averaged roughly 4,570 per month. In May 2026 alone the figure was 38,579 - more than eight times that monthly average.
Tech sector layoffs accelerated sharply in the first half of 2026, with AI the most-cited reason in three consecutive months.
Source: unsplash.com
What the Numbers Say
The profit paradox
The detail that makes this cycle different from past downsizing waves is the financial backdrop. Oracle cut 21,000 jobs while reporting $3.7 billion in quarterly net income - a 27% year-over-year increase. Microsoft committed more than $100 billion a year to AI capital expenditure while cutting 23,000 staff. Cisco posted record profits in the same period it removed 4,000 positions citing AI and security investment.
The usual logic of layoffs - a company is in trouble, it needs to cut costs - doesn't hold here. The companies doing the most cutting are reporting strong results. The cuts aren't defensive. They're structural: AI reduces the headcount needed to create a given unit of revenue, and the savings go straight into data center spend and GPU procurement. Amazon, Microsoft, Alphabet, and Meta together committed roughly $700 billion in capital expenditure for 2026 - nearly double their 2025 combined total - while simultaneously shrinking their workforces.
Challenger's own summary for May put it plainly: AI is "the most frequently cited layoff reason" for the first time in the report's history, and the figure "already far surpasses" the 2025 full-year total.
The roles being displaced
Customer support is the most documented category. Salesforce said publicly that AI agents now handle roughly half of its customer interactions, and the company has stopped backfilling support engineer roles as a result. Standard Chartered's framing - "lower-value human capital" replaced by machines - covers data entry, document processing, and compliance review functions.
Software engineering isn't immune. GitLab, restructuring for what it called the "agentic AI era," flattened up to three management layers and exited 22 countries, with about 350 staff affected. Block's Jack Dorsey was more direct: "Intelligence tools we're creating, paired with smaller teams, change what it means to build and run a company."
"Intelligence tools we're creating, paired with smaller teams, change what it means to build and run a company." - Jack Dorsey, Block
The acceleration curve
Challenger began tracking AI as a stated layoff reason in 2023. The annual totals tell a clean story:
- 2024: Baseline year, category new and small
- 2025: 54,836 cuts cite AI across the full year
- 2026 YTD (through May): 87,714 - already 60% above 2025's total
The trend line doesn't flatten in the Challenger data. May 2026 was the highest single month on record. If June runs at anything close to the May rate, the mid-year total will approach 130,000 - more than double the full-year 2025 figure.
Challenger, Gray & Christmas has tracked employer-stated layoff reasons since the 1990s. AI only entered the report in 2023.
Source: unsplash.com
What the Numbers Don't Say
Challenger's methodology has a built-in limitation that matters here: it captures self-reported attribution. Companies decide, usually in a press release or earnings call, how to frame a workforce reduction. That framing is then counted.
"Some companies cite AI as the reason for workforce reductions when the real drivers are overhiring, declining revenue, or cost-cutting pressure from investors." - industry observers quoted by multiple outlets tracking the 2026 layoff wave
A company that overbuilt its workforce during the 2020-2022 hiring boom now has a convenient and market-approved reason to cut. Saying "AI is enabling us to do more with less" produces a different investor reaction than "we hired too many people." Both statements can be simultaneously true, but Challenger's data can't distinguish between them.
What the data also doesn't capture is the silent version of the same shift: roles that simply aren't backfilled when someone leaves. Salesforce acknowledged it isn't replacing departing support engineers. That doesn't show up in cut counts. The 87,714 figure is probably an undercount of actual AI-related job reduction.
The reverse is also possible at the margins - companies where AI was genuinely not the driver but the press release said it was. Both distortions push in the same direction: the Challenger number should be read as a floor, not a ceiling, for how much AI is reshaping employment.
One more gap: the data is US-only. The firms cutting globally - Standard Chartered with 7,800 positions, Accenture across its consulting operations, HSBC - are making moves whose full workforce impact isn't reflected in an US-only tracker.
So What?
The Challenger report won't settle the debate over whether AI is the real driver of 2026 layoffs or a convenient label. It doesn't have to. What it shows is that employers are choosing to associate workforce reductions with AI at a rate that has no precedent in the data series, that the acceleration is still going up, and that the companies doing so are posting strong results while they do it.
For workers, the Salesforce and GitLab examples point to where the exposure concentrates: support functions, document processing, compliance review, and management layers in roles where AI tooling has become good enough to do the work. For investors, the pattern is fairly legible - inference infrastructure spending is the other side of this ledger, and it's growing at the same rate the headcount is shrinking.
The question that the Challenger data can't answer is how much productive capacity gets written off when a company cuts 4,000 people and replaces their output with an AI system that costs a fraction of the wage bill but fails in ways the humans didn't. We'll find out in the earnings calls.
Sources:
