This article was produced by the AETW editorial team from the AETW AI layoffs tracker dataset.

Summary

In AETW's layoffs tracker, AI was cited in layoffs affecting 92,913 of 129,936 workers from January to September 2026, or 72%, up from 17% in the same months of 2025. Six companies account for 69% of those AI-cited cuts, and AI-cited layoffs fell to almost nothing after June.

AI layoffs in 2026: the headline number

In AETW's AI layoffs tracker, companies or credible reporting tied 92,913 of the 129,936 tracked job cuts from January 1 through September 23, 2026 to AI. That is 72% of laid-off workers. In the same months of 2025, the share was 17%: 15,665 of 92,882.

Bar chart: AI was cited in layoffs affecting 17% of tracked workers in January to September 2025 and 72% in January to September 2026
Source: AETW AI layoffs tracker, events with a reported headcount, January 1 to September 23 of each year. Data as of September 24, 2026.

Counting events instead of workers tells the same story at a smaller scale. AI was cited in 114 of 347 layoff events in 2026 (33%), against 38 of 262 in January to September 2025 (15%). The worker share is higher because the biggest cuts of the year were the ones most likely to mention AI.

Put simply: in 2026, about seven in ten workers who lost their jobs in a tracked layoff were cut by a company that pointed to AI, or whose cuts credible reporting tied to AI. The US carries most of it. US companies account for 81,724 of the 92,913 AI-cited cuts, or 88%.

Two big waves, then AI-cited cuts nearly disappeared

AI-cited cuts came in two waves: January to March, led by Amazon and Dell, and May to June, led by Oracle, Meta, Cisco and Intuit. May and June alone account for 45,895 AI-cited job cuts, about half the year's total.

Stacked column chart of 2026 tracked layoffs by month, split into AI-cited and other reasons; AI-cited cuts are large from January to June and almost zero from July to September
Source: AETW AI layoffs tracker, events with a reported headcount. September covers September 1 to 23, 2026.

After June, AI-cited cuts nearly disappeared from the tracker. From July 1 to September 23, it logged 2,636 AI-cited job cuts, against 90,277 in the first half. Layoffs themselves continued: July still had 11,000 tracked cuts, and only 1,763 of them were tied to AI.

We don't know why yet, and the data can't tell us. One plausible reason is legal risk: in July, 26 current and former Meta employees sued over AI layoffs they say penalized workers on medical, parental and family leave, and naming AI in a layoff memo started to look like a liability. Another is that the biggest AI restructurings were planned for the first half and are simply done. We will watch whether the pattern holds in the October and November data.

A handful of giants drive the AI layoff number

Six companies account for 64,000 of the 92,913 AI-cited job cuts, or 69%. Oracle alone is 21,000, after its annual filing tied a smaller workforce to AI adoption. Amazon's January corporate cuts add 16,000, and Dell's filing showed 11,000 fewer workers.

Amazon is the most contested row. The tracker counts it as AI-cited, but Amazon framed the January cuts as removing layers and bureaucracy, and Andy Jassy said in 2025 that the cuts were "not even really AI-driven." Leave Amazon out and the 2026 share is 59%, still more than three times 2025's.

Table of the six largest AI-cited layoffs in 2026: Oracle 21,000, Amazon 16,000, Dell 11,000, Meta 8,000, Cisco 4,000 and Block 4,000
Source: AETW AI layoffs tracker, with each row linked to the company statement, filing or news report. Data as of September 24, 2026.

A few of the largest US tech employers account for most of the AI-cited total, so the 72% figure says more about how big companies framed big cuts than about how many companies are replacing staff with AI. Of the 114 AI-cited events in 2026, only 83 came with a reported headcount, and 67 of those 83 cut fewer than 1,000 people.

How many jobs has AI replaced? What this data can and can't say

This report counts layoffs where AI was cited, not jobs AI replaced. A company can name AI to explain a cut it would have made anyway, and AI can cost jobs through hiring freezes that never show up as a layoff. For the hiring side, Stanford Digital Economy Lab's research on workers aged 22 to 25 in AI-exposed jobs is the better guide.

How we built the numbers:

  • Source: the AETW AI layoffs tracker, a public CSV of reported layoff events with a row-level source link for each.
  • Period: January 1 to September 23, 2026, compared with January 1 to September 30, 2025. The dataset was last updated September 24, 2026.
  • AI-cited means a company statement, filing or credible reporting tied the event to AI, automation or a related skills shift. Every 2025 and 2026 row carries a Yes or No flag.
  • Worker totals use only events with a reported headcount: 223 of 347 events in 2026. Events with only a percentage are left out of the sums.
  • Industry labels are our editorial classification. Counts are reported figures and may be approximate.

Use the data

The full dataset is free to download and reuse with a link back to the tracker. Reporters and researchers can cite this report as "AETW AI Layoffs Report, November 2026" and should include the data date, September 24, 2026. We publish the next report in early December, with October and November events added.

Sources