AI Layoffs 2026: What 392 Tech Layoffs Actually Said About AI

Are Companies Really Laying Off Workers Because of AI?

Announcements can't prove cause, but they show what companies chose to say, and few said AI cuts, or will cut, their staff needs. Of the 258 tech layoff events in 2026 with a company-stated reason, about 1 in 8 said so, more said they were refocusing on AI, and most gave a reason with no AI link. Another 134 events had no company reason in the source we read.

What Companies Actually Said When They Cut Jobs

"AI cuts staff needs" means the company said AI reduces, or will reduce, how many people it needs, in statements like Block CFO Amrita Ahuja's "smaller, highly talented teams using AI to automate more work." "Refocusing on AI" covers lines like Envato CEO Hichame Assi's "we are reducing operational costs to ensure we can invest further in product, marketing and AI."

9 events blamed AI for hurting the business itself. At Pipedrive, the sales CRM, the CEO said, as the source reports it, that "more small businesses are asking artificial intelligence (AI) for software" recommendations.

8 denied that AI replacement or AI efficiencies drove the cuts, like HubSpot, whose CEO Yamini Rangan wrote that its October cut "is not driven by AI-related efficiencies." Canva's November 3 announcement, which affected 1,271 people through departures, role changes and relocations, gave a business reason: growth "has slowed for the first time in a decade."

Inside the "AI Layoffs" Label

layoffs.fyi tags a layoff as AI when the company or the press attributed it to AI. That's a fair definition for a tracker, and it means the tag mixes two different things.

Amazon's 16,000-person cut in January carries the AI tag, but Amazon's own memo talked about "reducing layers." This comparison doesn't establish whether the tag is right or wrong.

Headlines show the same gap. In the 61 events where the article framed the cut as AI-driven in the reporter's own words, 17 came with a company claim that AI cuts staff needs, and 7 had no company reason at all.

Company Statements Linking Cuts to AI Peaked in May

Challenger, Gray & Christmas, which counts announced US job cuts rather than events, separately reported 1,265 AI-cited cuts in October, "its lowest monthly total this year." It also said market and economic conditions overtook AI as the top reason for 2026.

Every 2026 Layoff Where the Company Mentioned AI

These are the 99 layoff events where a statement by or attributed to the company brought up AI, newest first. Each excerpt is copied from the linked source, and some are the reporter's summary of what the company said.

No layoffs match that search.

Who's Still Hiring After AI Layoffs

On November 5, a script searched for public job boards for all 346 companies and retrieved 12,869 open postings from the 119 boards it could match to a company with at least one opening.

80 of those 119 companies (67%) were advertising at least one marketing role, including 10 of the 11 that said AI cuts staff needs. AI-focused roles made up 7.4% of postings, and marketing roles made up 3.6%.

A posting isn't a hire, and a board can include roles in many countries. "AI-focused" means Claude judged the job title to name AI, machine learning or a related field (LLMs, generative AI, AI agents, computer vision, NLP) as the role's focus.

What This Means for Marketers

1. Read the announcement, not the headline: the company's statement usually says whether AI cut its staff needs, the company was refocusing on AI, AI hurt the business, or the company gave a reason with no AI link. Each one points to a different risk for your job.
2. Check current openings: on November 5, about two-thirds of the 119 companies with a matched job board and open postings were advertising marketing roles, and the companies left out may differ. My study of 1,229 marketing job posts shows what those roles ask for, and my list of AI job boards is a good place to look.
3. Show what you ship with AI: AI-focused roles outnumbered marketing roles on these boards. My advice is to keep a record of real work you finished faster with AI tools, because that's easier to point to than a list of tools.

How This Field Test Was Built

• Why "field test": the analysis rules were written down before labeling and before the job-board pull. The headline test compared press framing with company statements and needed a clear result under a worst case where every missing company statement counted as an AI claim. It didn't get one (17 of 61, rising to 24 of 61 in the worst case), so the findings are descriptive. The hiring check also fell short of its targets of 120 companies and 30 that tied cuts to AI (119 and 25).
• The list: every layoff event on layoffs.fyi dated January 1 to November 4, 2026, pulled November 5 (392 events at 346 companies). layoffs.fyi says its data is free to use with attribution, and this page uses its list and AI tag, not its job counts.
• The sources: a script saved the article layoffs.fyi links for each event. 78 thin sources (paywalls, cookie walls, empty pages) each got one follow-up Google search, and any alternate story found was shown to both coders. After that, 76 events still had no usable source text.
• The coding: Claude Sonnet and GPT (via Codex) coded every event from the full saved text with the same written rules, blind to each other and to the layoffs.fyi tag. Only company statements or reporting explicitly attributed to the company counted, never anonymous sources, analysts or the reporter's own framing.
• Rule changes, dated: a 40-event pilot came in at 87.5% agreement, under the 90% bar, so the rules were clarified and checked on 40 fresh events (95.0%). A fact-check of the first full run then found that companies blaming AI for lost business had no category of their own, so a final version added "AI hurt the business," renamed two categories to match what they cover, and both coders re-coded all 392 events.
• Agreement: on the final rules the coders agreed on the reason for 368/392 events (93.9%) and on press framing for 95.9%. A separate Claude model (Opus) settled the 24 reason disagreements against the source and logged each call. A script matched every excerpt to the saved source, ignoring differences in case, punctuation and spacing.
• The hiring check: on November 5, a script retrieved open postings from the public Greenhouse, Lever, Ashby, SmartRecruiters, Workday and Oracle boards it could match to a company (Oracle's board reported 2,800 jobs and 2,798 came back), and 16 matches pointing to the wrong company or a side account were rejected. Boards with no open postings couldn't be verified and were left out. Claude coded job titles, and GPT agreed on 98.0% of AI-role labels and 99.5% of marketing-role labels in a 200-title check. Big employers whose boards the script couldn't retrieve, like Amazon and Microsoft, aren't in the hiring numbers.
• Limits: announcements show what companies chose to say. A "no reason given" event may have a reason in a source we didn't read, and whether marketing roles were cut stayed unclassified for 309 of 392 events (mostly because the source didn't name the teams, plus cases where the two coders disagreed), so this data can't count marketing jobs lost.
• Other trackers: TechCrunch, ailayoffs.org and ailayofftracker.com track layoffs where AI was cited, and the last two separate company statements from press reports. This page adds the full list of 2026 tech layoff events, so the AI cases read as a share of all of them, plus a per-company hiring check.
• Who did the work and disclosure: scripts collected the data, Claude and GPT coded and checked it, and I chose the angle and wrote the takeaways. I teach marketers to use AI, sell Claude Code products, and have written posts about using AI to do work a team or agency would do, so weigh my takeaways with that in mind.

Doser, R. (2026). AI layoffs 2026: what 392 tech layoffs actually said about AI. ryandoser.com. https://ryandoser.com/research/ai-layoffs-2026/