It’s Not a Warning Anymore: The AI Job Cuts Are Here

I started Durable Earnings with a simple premise: AI was going to change the math on millions of careers, and most people wouldn’t see it coming until after it arrived.

I wanted to be wrong about the speed.

I wasn’t.

On May 20th, Meta cut 8,000 jobs and forcibly reassigned another 7,000 employees into AI-focused teams. That’s roughly 20 percent of the company reshaped in a single week. The reassigned workers weren’t asked. A Meta VP wrote in an internal memo that the transfers “aren’t optional.” Both new units — Applied AI Engineering and the Agent Transformation Accelerator — exist for one purpose: building AI agents that can automate the work humans currently do.

Read that again. Meta is reassigning thousands of its own people to build the systems that will eliminate the need for people like them. That’s not speculation about the future. That’s an internal memo from last Tuesday.

And Meta is not alone.

The Numbers Have Stopped Being Abstract

Over 92,000 tech jobs have been eliminated in the first five months of 2026. That number comes from industry tracking firms, not opinion columns. April was the worst month for tech layoffs in two years.

Here’s a partial list of what the first half of this year looks like:

Meta — 8,000 jobs cut, 7,000 reassigned to AI, more cuts signaled for August and fall. Capital expenditure guidance raised to $125–145 billion, nearly double 2025. The money that used to pay salaries is now buying GPUs.

Amazon — 16,000 corporate roles eliminated in January alone, on top of 14,000 cut in October. The stated reason: reducing bureaucracy. The unstated math: AI handles coordination cheaper than middle managers do.

Dell — 11,000 jobs cut across traditional PC and server divisions. AI server revenue up 40 percent. The old business funds the new one. The people in the old business fund their own replacement.

Cisco — 4,000 jobs to refocus on AI. Quarterly results strong. Stock fine. The layoffs aren’t about survival. They’re about acceleration.

Atlassian — 1,600 employees, 10 percent of the workforce. The CEO said it plainly: “It would be disingenuous to pretend AI doesn’t change the mix of skills we need or the number of roles required.”

Oracle — Reports of up to 30,000 positions eliminated.

Coinbase — 700 jobs, 14 percent of total workforce. The CEO said the company needs to become “AI-native.”

Standard Chartered — Plans to eliminate 8,000 support roles over four years.

And this is only what’s been announced publicly. Every major consultancy I’ve spoken with — and I’ve spent my career inside this world — is running internal models on how many roles AI makes redundant. Most of those numbers haven’t been shared with the people who occupy the roles.

What Changed: It’s Not the Technology. It’s the Economics.

A year ago, most of these companies were still talking about AI as a productivity tool. Something that would help existing teams work faster. An augmentation story.

That story is over.

What happened is simple: the economics tipped. When an AI agent can handle a task that previously required a $120,000-a-year employee, the CFO doesn’t need a philosophy seminar on human-AI collaboration. The math is the math. And companies don’t ignore math — not at this scale, not when competitors are already moving.

Meta’s projected capital expenditure for 2026 is between $125 and $145 billion. That’s more than twice what they spent last year and nearly four times 2024. Where does that money come from? Partly from the salaries they’re no longer paying. The people being let go aren’t being replaced. They’re being converted into infrastructure spend.

This is the mechanism I described when I launched this site. Not a dramatic overnight collapse. Not Hollywood robots walking into offices. Just a quiet, quarter-by-quarter reallocation: human payroll into AI infrastructure. It looks boring from the outside. It’s devastating from the inside.

Who This Hits Hardest — And Why Our Readers Should Pay Attention

If you’re reading Durable Earnings, you’re probably between 40 and 60. You’ve built a career on knowledge work — analysis, coordination, reporting, project management, operations, financial planning, HR processes, sales support. Structured cognitive work. The kind that requires intelligence, experience, and discipline.

That description now overlaps almost perfectly with what AI agents are being built to do.

The Meta layoffs weren’t concentrated in the mail room. Engineering and product divisions bore a disproportionate share. The Amazon cuts hit corporate roles — the planning, logistics, and coordination functions that used to be the backbone of white-collar employment. Dell cut traditional divisions, not the AI side. The pattern is consistent: legacy knowledge work shrinks, AI investment grows, and the people in between absorb the cost.

The uncomfortable truth for our age group specifically: we’re expensive. A 50-year-old with 25 years of domain expertise commands a significant salary, benefits package, and institutional knowledge premium. That premium used to be the reason companies kept us. Now it’s the reason the ROI on replacing our output with AI looks so attractive.

This is not about whether you’re good at your job. Most of the people being laid off are good at their jobs. It’s about whether the category of work you do is becoming cheaper to produce by other means.

What I Got Wrong — and What I Got Right

When I wrote the letter that became the foundation of this site, I laid out a scenario where companies would quietly discover they could operate with far fewer knowledge workers, and once one competitor cut costs, the rest would follow. I described layoffs arriving in waves that feel temporary at first — explained as restructuring, efficiency, or market conditions.

I thought that process would unfold over two to three years.

It took about eight months.

The World Economic Forum surveyed companies last year and found that 41 percent expected to reduce their workforces within five years because of AI. That number felt aggressive when it was published. It now looks conservative. We’re watching companies cut 10 to 20 percent of their workforce in single announcements, not over five-year horizons.

What I got right was the mechanism: this isn’t a recession-driven downturn. This is a structural reallocation. The jobs aren’t coming back when the economy improves, because the economy is improving by eliminating them. Meta’s stock is fine. Cisco’s quarterly results are strong. Dell’s AI revenue is surging. The companies doing the cutting are not struggling. They’re optimizing.

That distinction matters more than anything else in this article. If you’re waiting for a recovery that restores the old structure, you’ll be waiting for something that isn’t arriving.

This Isn’t Doom — But It Is Urgency

I’ve said from the beginning that Durable Earnings isn’t about fear. I believe there’s a positive path through this — a future where disruption leads to transformation rather than collapse. I still believe that.

But I also believe the people who navigate it best will be the ones who saw it coming and moved deliberately. That window is narrower than it was six months ago.

Everything on this site — the guides to income diversification, the honest assessments of which roles are at risk, the practical steps for building income that doesn’t depend on a single employer — was built for exactly this moment. Not a theoretical someday. This week. These headlines. These numbers.

If you’ve been reading and thinking about acting, the data is now saying what I was saying before the data existed: the time to diversify your income, audit your career exposure, and build something durable is before the restructuring memo lands in your inbox. Not after.

What You Can Do This Week

I’m not going to pretend one article fixes the problem. But here are three things worth doing before Monday:

Audit your task portfolio. Write down what you actually do in a given week. Not your job title — your tasks. How many of them involve processing information versus making judgment calls that require relationships, context, or physical presence? If more than half of your week is structured information work, your role has meaningful AI exposure. That’s not a verdict. It’s a starting point.

Start one income stream that isn’t your employer. It doesn’t have to be dramatic. A freelance skill. A small digital product. A consulting relationship. Something that generates revenue independently. The goal isn’t to replace your salary tomorrow. It’s to have something already moving if the salary changes.

Talk to your employer honestly. Not from a place of fear — from a place of strategy. Ask what the company’s AI roadmap looks like for your department. Ask where they see your role in 18 months. If the answer is vague, that tells you something. If the answer is specific and includes you, that tells you something better.

None of this is panic. All of it is preparation. And preparation is what separates the people who navigate disruption from the people who are caught by it.

Sources

Yahoo Finance, “Meta Layoffs 2026: 8,000 Jobs Cut as $145B Goes to AI,” May 20, 2026
TechJournal, “Meta Layoffs Begin Today: 8,000 Jobs Cut as $145B Goes to AI,” May 20, 2026
The Register, “Those spared latest Meta job cuts forcibly reassigned to AI roles,” May 20, 2026
Intellectia.ai, “AI Layoffs 2026: Tech Workforce Crisis and Investment Outlook,” May 21, 2026
eWeek/Tech Insider, “150K+ Tech Jobs Cut in 2026 — Who’s Next?” April 2026
Programs.com, “List of Companies Announcing AI-Driven Layoffs,” May 2026
Intellizence Layoff Tracker, “Companies that announced Major Layoffs,” updated May 20, 2026
Business Insider, “Companies laying off staff this year,” updated May 2026

—

Continue reading: Your Income in the AI Era — the pillar guide to understanding how AI reshapes careers, income, and what comes next.

This article is for informational purposes only and is not financial advice. Consult a qualified professional for personalized guidance.

Continue reading: Read the pillar — Your Income in the AI Era