The Q1 2026 tech layoff tallies are in, and they read like a eulogy written by a productivity consultant: 78,557 tech workers laid off in the first three months of the year, and 47.9% of those cuts — 37,638 humans — explicitly attributed to AI and workflow automation, according to tracking reports compiled this week.
The ‘AI Made Me Do It’ Defense
At least eight companies announced AI-related layoffs of 10,000+ employees each this quarter: Accenture, Amazon, Citigroup, Dell, Intel, Microsoft, TCS, and UPS. That is not a list of struggling startups. That is the list of places your uncle told you to apply.
March alone saw 60,620 cuts, with AI cited as the top reason in Challenger’s monthly report. Customer support, coding, and data processing took the heaviest hits — which, to be fair, are precisely the tasks AI is genuinely getting good at. It’s not nothing.
But Is It Really AI, Or Just… Vibes?
Here’s where it gets interesting. Harvard Business Review ran a piece this quarter arguing companies are laying off workers because of AI’s potential — not its performance. Translation: we haven’t necessarily proven the AI can do the job, but the CFO read a McKinsey deck and now we’re doing this.
The blockchain-adjacent commentariat has been even blunter, accusing executives of using “AI” as a rhetorical fig leaf for garden-variety cost-cutting. When you can’t blame the economy and you’ve already blamed interest rates, blaming the large language model is the 2026 equivalent of “market conditions.”
What This Means For Actual Humans
About 20% of full-time US employees now report that AI has replaced some of their work, per NBC’s survey. That doesn’t mean 20% of workers are unemployed — it means 20% of workers are watching tasks evaporate from their job description in real time, which is its own kind of slow-motion layoff.
The workers that have been cut say they weren’t prepared. Nobody is. The people writing the onboarding materials were laid off in Q1.
The Part Where We Try To Be Honest
Some of these cuts are real automation doing real work. Customer service bots now resolve 70–80% of routine inquiries. AI coding assistants handle the boilerplate that used to keep three junior engineers billable. That’s not cope; that’s observable.
The rest is something harder to name — a kind of pre-emptive surrender, where middle management decides the AI will be able to do it soon enough, so why keep paying someone in the meantime. The paycheck stops before the automation arrives. That’s the gap workers are falling into.
The AI did not write this article. But ask again next quarter.
Sources: