More than 200,000 tech layoffs in 2026: AI is the most cited excuse, but not the only reason for the cuts

By Quartz (with data from the Yahoo Tech tracker) · November 5, 2026.

The tech sector has racked up more than 200,000 layoffs in 2026 with two months still to go, according to the TrueUp tally cited by Quartz. In all of 2025 there were more than 245,000, so this year's pace appears higher. Quartz groups entries from the Yahoo Tech tracker into ten sectors, and makes clear that it is an editorial selection: it does not total or rank sectors by volume. Several figures are announcements or third-party reports, and some companies have not disclosed how many people are affected.

The largest cases, according to the data compiled in the article: Meta is cutting 10% of its workforce (about 8,000 jobs) and closing 6,000 open positions; Microsoft announced 4,800 layoffs in July and a voluntary buyout program; Oracle shed 21,000 jobs in a year (13% of its workforce, according to a June 23 filing); Amazon is at around 30,000 between October and January; Xbox expects 3,200 departures through fiscal 2027. Uber (3,300), Intuit (3,000), Visa (2,600) and Netflix (about 1,000) round out a picture that is no longer limited to engineers: it also hits recruiters, lawyers, product managers and HR.

The key nuance is the stated reason. In enterprise software and cloud, Quartz counts seven out of ten companies linking the cuts to AI. The most explicit case is Atlassian: its CEO, Mike Cannon-Brookes, said the cuts (about 1,600 jobs, or 10%) would free up funds for more investment in AI. Oracle, according to reports, is seeking to offset spending on chips and data centers. But in video games, none of the six companies cited blames AI: Epic, with more than 1,000 layoffs, attributes them to Fortnite's decline, and its CEO, Tim Sweeney, said explicitly that AI played no role. In marketplaces, Uber, Etsy, Patreon and eBay talk about reorganizing teams, not automating. Patreon even states that it is not cutting "because it believes AI replaces people".

Our take: the easy headline, "AI is destroying jobs", poorly describes what is happening in 2026. What the overall picture shows is something more mundane and at the same time more troubling: AI is competing with salaries for the same budget. Companies are cutting payroll to pay for chips, data centers and AI platforms, even though the software has not yet replaced that work. It is an anticipatory substitution of capital for labor, made in advance on the promise of productivity that has yet to be proven.

There is also a structural pattern worth separating from AI: flatter organizations (Uber is removing middle layers and shrinking micro-teams by about 50%; Canva wants to reduce "over-specialization"; Wix wants to be "flatter"). It is the "small teams" logic that already existed, and AI serves as both alibi and tool. Corporate narratives should also be read with caution: "we're doing it because of AI" plays well with investors, and "it has nothing to do with AI" plays well with employees and regulators. Both are talking points; the hard data are the headcount figures.

Overall, the transition we have been anticipating is real and harsh, and 2026 is making it visible: the pressure falls on support roles (HR, legal, recruiting) and middle layers, and the demand by more than 4,500 Google employees for buyouts and guaranteed severance shows that social bargaining is already starting. Governments such as California's developing tracking tools is a reasonable step: without measurement there can be no evidence-based public policy.

The long term is still what matters. If today's AI investment delivers on its promise (scientific discovery, medicine, productivity that makes goods and services cheaper), the resulting abundance could free people from jobs no one will miss. But that transition is neither automatic nor painless: it depends on the costs of adjustment being shared more fairly than in 2026, when those footing the bill for the bet are today's workers. What to watch: how many of these companies show in 2027 that productivity per employee is genuinely rising, and how many have simply cut.