Big Tech Is Laying Off 60,000 People and Calling It AI Investment. Here's What's Actually Happening.

More than 60,000 tech workers have been laid off in 2026. We're not even through Q1.

The announcements follow a pattern. Amazon cuts 16,000 jobs while investing billions in AI infrastructure. Block — the Jack Dorsey payments company — eliminated 40% of its entire workforce in a single move. That's 4,000 people, gone. Meta cut 1,500 employees from its Reality Labs division. Ericsson, ASML, semiconductor companies across the sector — thousands more.

Here's the line that keeps appearing in every press release: "We're restructuring to invest in AI."

Let's talk about what that actually means.

The Numbers Don't Lie

In 2025, AI was cited as a direct cause in fewer than 8% of tech layoff announcements. Through early March 2026, that number has jumped to 20.4%. One in five layoffs is now being explicitly attributed to AI replacing human work — and that's only what companies are openly admitting.

That's roughly 9,200 out of 45,000 confirmed layoffs where the company looked workers in the face and said: a machine is doing your job now.

Analysts project the sector is on pace for over 264,000 total layoffs in 2026 if current trends hold. That would exceed all of 2025, which was already brutal at nearly 246,000.

What "Restructuring for AI" Actually Translates To

Amazon's $100 billion AI investment is going primarily into infrastructure — data centers, chips, cloud computing. That creates some construction and technical operations roles. It does not replace 16,000 customer service agents, content reviewers, and middle managers.

Block's 40% cut was more blunt. Jack Dorsey told employees directly that the company had grown too bureaucratic, and that AI tools could handle work that previously required large teams. Four thousand jobs, one announcement, done.

The translation: work that humans used to do can now be done faster and cheaper by AI. Some different humans — prompt engineers, ML researchers, AI product managers — are being hired. But not at the same volume as those being cut.

The Argument That AI Creates More Jobs

This argument is real and historically grounded. The printing press eliminated some roles and created far more. Computers did the same. Automation in manufacturing eventually created more jobs than it destroyed.

But two things are different this time.

First: the pace. Previous technological disruptions played out over decades. Workers had time — not always enough, but some — to retrain and adapt. This disruption is happening in years, not decades. The runway is shorter.

Second: the breadth. Past automation primarily hit routine manual labor and clerical work. AI is hitting cognitive work: writing, coding, legal research, analysis, customer service, financial modeling. The assumption that "knowledge workers are safe" is being tested right now, not in the abstract.

"AI will eventually create more jobs than it destroys" is a long-game argument. The people losing jobs right now are playing a short game.

Who Is Actually Getting Hired

The companies cutting most aggressively are also the companies building the most AI capability. That's not coincidence — it's conversion. Human operational capacity is being replaced by AI operational capacity.

The roles that are growing: AI safety and alignment engineers, data infrastructure, AI product managers, machine learning researchers. These are high-skill, high-barrier roles that require years of specific technical experience.

The math doesn't balance for someone who was a content reviewer at Amazon or a customer support rep at Block. "Learn to code" is the punchline of 2012. The actual advice is more specific and harder: you need to understand how AI systems work, where they fail, what they can't do, and how to build products and processes around them.

Q2 and Q3 earnings calls will be worth watching. Every major tech company will simultaneously report AI revenue growth and "workforce efficiency gains." The numbers will be right there side by side.

What It Means for Your Career

If you're in tech and haven't spent serious time with AI tools — not just using them casually, but understanding what they can and can't do, where they're unreliable, how they're being deployed in your industry — the next 18 months are the window to do that.

This isn't alarmism. Current models have real limitations: hallucinations, context failures, inability to handle novel situations well, compliance and liability risk in regulated industries. There's still a significant role for human judgment in most knowledge work.

But the companies are restructuring. They're telling you what they value. The question is whether you're building skills that fit the next three years — or still doing the work a language model is already handling in the job posting you'll apply for next.

The tech layoff wave of 2026 isn't a blip. It's a signal. And the companies sending it are being more honest about the reason than they usually are.