Jack Dorsey didn't apologize. He explained. Block — the fintech behind Square and Cash App — laid off 4,000 employees in early 2026, wiping out 40% of the company's entire workforce. His statement was blunt: these cuts exist so the company can "move faster with smaller, highly talented teams using AI to automate more work." No euphemisms. No corporate misdirection. Just the math.

That kind of candor is spreading. And if you work in tech, it should focus your attention.

As of late March 2026, 59,121 tech workers have received layoff notices this year — an average of 704 per day, every single day. Research firm RationalFX tracked the source of those cuts and found that more than 9,200 — roughly 1 in 5 — were directly attributable to AI adoption and automation. Not a market correction. Not overhiring. AI, specifically and explicitly.

The breakdown is precise enough to be unsettling. Amazon cut 2,700 positions from its cloud operations team the same week it announced a new AI-powered infrastructure management system. Google eliminated 1,200 roles in its advertising division and replaced them with what it publicly calls "autonomous campaign optimization." Microsoft let go of 800 program managers while simultaneously expanding its Copilot agent workforce. Meta is reportedly considering cuts of up to 15,000 workers — roughly 20% of its total headcount — while committing $115 to $135 billion in AI infrastructure spending in 2026 alone.

These aren't parallel events. These are cause and effect, stated openly in investor calls and press releases.

The roles disappearing follow a clear pattern: software operations, customer support, logistics planning, financial modeling, content moderation, marketing analytics. These are exactly the categories where AI agents have become competent enough to replace, not just assist. If the current pace of cuts holds, industry trackers project 265,000 tech job losses by December. That would make 2026 the most damaging year for tech employment since the dot-com era — except these roles aren't returning in the next cycle. They're being permanently replaced.

The U.S. accounts for roughly 80% of the global total, meaning this is concentrated, not diffuse. And across industries, 52% of layoff announcements are now explicitly linking cuts to AI and automation efficiencies.

My Opinion

Here's what bugs me: we've been watching this unfold for two years, and the narrative hasn't caught up. The official story is still "AI augments workers." There's a version of that which is true — some jobs are genuinely becoming more powerful with AI. But when Jack Dorsey explicitly cites AI as the reason 4,000 people no longer have jobs, when Amazon eliminates an entire cloud ops team and immediately replaces the function with software, "augmentation" is the wrong word. That's replacement. Call it what it is.

What's changed in 2026 is that companies have stopped hiding it. They're telling shareholders directly: AI allows us to do more with fewer people, and the stock price responds positively. This is now a feature, not a side effect. The executives who are most honest about it — Dorsey being the clearest example — aren't apologizing; they're describing strategy.

I'll be direct: if you're in a role that involves repeatable cognitive tasks — even complex ones — the question isn't whether AI will affect your position. The question is when, and whether you'll have prepared an answer. "I'm good at my job" is not a moat anymore. What you need is irreplaceability: judgment, relationships, domain expertise that hasn't been encoded into a model yet, the ability to manage the AI itself. The 59,000 who lost jobs this year didn't all see it coming. The next 59,000 have no excuse not to.

Author: Yahor Kamarou (Mark) / www.humai.blog / 28 Mar 2026