Meta Is Cutting 8,000 Jobs Because of AI, and This Time It Feels Different

Eight thousand people. That is the first wave.

Meta plans to cut about 10% of its global workforce on May 20, according to Reuters, with more layoffs coming in the second half of the year. Executives are watching AI capabilities and may adjust the size of the second round based on how fast the technology advances. Let that sink in. The next round of job losses depends on how quickly AI replaces what those workers do.

This is not Meta's first layoff rodeo. In late 2022 and early 2023, the company cut about 21,000 jobs during what Zuckerberg called the "year of efficiency." Back then, the stock was in freefall, Meta was over-hiring after COVID boom assumptions proved wildly optimistic, and the cuts felt like a reckoning. It was uncomfortable, but it made sense in the way corporate restructurings always make sense: things got bad, something had to give.

This time is different. That is the part I keep turning over.

The Financial Picture Makes No Sense for a Layoff Story

Meta generated over $200 billion in revenue last year and posted a $60 billion profit. The stock is up 3.68% since January. The company employs nearly 79,000 people and is committing up to $135 billion in capital expenditure this year alone, with a $600 billion AI infrastructure pledge through 2028. Layoffs.fyi has tracked 73,212 tech workers losing their jobs so far this year, and for all of 2024 the total was 153,000. Those numbers are not small.

And yet here is Meta, a company that is printing money and spending more than most countries' GDPs on AI infrastructure, cutting jobs. Not because it has to. Because it wants to.

Amazon has trimmed 30,000 corporate employees in recent months, nearly 10% of its white-collar workforce. Block, the fintech company, chopped nearly half its staff. In both cases, executives tied the cuts directly to artificial intelligence. The language is identical: efficiency, restructuring, AI-driven productivity gains.

I have been watching tech layoffs for years. I have never seen a cycle where the companies making the most money are also the ones cutting the most people.

What "Applied AI" Actually Means

In recent weeks, Meta reorganized teams across Reality Labs and transferred engineers throughout the company into a new organization called "Applied AI." The mandate is explicit: build AI agents that can write code and carry out complex tasks autonomously. Zuckerberg is not hiding this. He is broadcasting it.

The company also recently built manager-to-employee ratios of up to 1:50 in parts of the new AI engineering organization. The old standard was closer to 1:15 or 1:10 in typical tech companies. Zuckerberg has spoken openly about flattening the company's structure. What that means in practice is fewer bosses and more AI. The middle layer, the one that coordinates and translates and manages people, is the layer getting removed.

Sales, recruiting, and global operations were trimmed in March. Reality Labs lost over 1,000 people in January. Middle management is next, if not already included in that 8,000.

Here is the uncomfortable part: this is not a failed company cutting fat. This is a company so confident in AI replacing work that it is cutting while profitable. The normal rules do not apply.

The Numbers Should Worry More People Than They Do

Let me put this in perspective. Meta is cutting 8,000 people in one wave. There are more waves coming. Layoffs.fyi data shows 73,212 tech workers have been cut already this year, tracking toward matching or exceeding 2024's 153,000.

Amazon cut 30,000. Block cut nearly half its entire staff.

Block is the one that sticks with me. Nearly half. Not because the company was failing but because AI made nearly half the work disappear. That is not a restructuring. That is a rewriting of what a company needs to employ.

DeepSeek, the Chinese AI startup, is simultaneously raising $300 million at a $10 billion valuation. AI talent wars are still global and funding is still flowing even as Big Tech cuts headcount. The companies that are winning the AI race are also the ones racing to need fewer humans. That contradiction is not getting resolved in this cycle.

I do not know how to feel about this. I genuinely do not. The productivity argument is real. If AI agents can handle routine coordination work, decision-making within defined parameters, and code execution at scale, then the productivity gains are real and the economic logic is not wrong. But "the economic logic is not wrong" is a cold way to describe 8,000 families finding out their employer decided they are subtractable.

What Makes This Round Different

The 2022-2023 layoffs were about correction. Companies over-hired, growth assumptions collapsed, and the market forced a reckoning. Uncomfortable, but rational in the way a company being forced to resize is always rational.

These layoffs are about substitution. The company is not cutting because it has to. It is cutting because it believes AI can do the work more efficiently, and the financial position gives it the runway to make that bet without the market forcing its hand.

That is a different kind of confidence. That is the confidence that comes from having $60 billion in profit and believing the next $60 billion does not require the same headcount.

And the second wave, the one that depends on how fast AI capabilities advance, is the part that makes this structural. This is not a one-time correction. This is a standing policy. If the AI gets better, more people leave. If the AI gets much better, more leave. The relationship between workforce size and AI capability is now explicit corporate planning, not just market pressure.

The Productivity Argument Is Not Wrong, But It Is Not Complete Either

I want to be honest about this, because I see both sides clearly and they are both true.

AI-assisted workers can be more productive. Fewer managers coordinating fewer people through a flatter structure, with AI agents handling execution and coordination, might genuinely produce better outcomes at lower cost. The companies making this bet are not stupid. They are reading the same data I am and making the same bet I would make in their position.

But "better outcomes at lower cost" is an aggregate claim. It does not tell you who bears the cost of the transition. The companies and their shareholders benefit. The laid-off workers and their communities absorb the loss. Those are not the same people.

And the pace is not slowing. 73,000 jobs this year. More waves coming. The next round at Meta depends on AI capability improvements. That is not a company preparing for a transition. That is a company running the transition as fast as the technology allows.

I do not have a clean answer for what the alternative is. Asking Meta to not cut jobs when AI makes those jobs unnecessary is asking a company to be something other than a company. But the outcome of that logic, applied across the entire economy, is not a productivity story. It is a distribution story. And distribution stories are where political economy lives.

What This Means for the Rest of Us

If you work in tech, in operations, in anything involving coordination and execution that a sufficiently capable AI agent can handle, this is your context. Meta is not an outlier. It is a leading indicator. The company with the most money, the most data, and the most aggressive AI buildout is explicitly replacing human labor with AI agents and calling it efficiency.

Amazon did the same thing. Block did it more aggressively. The fintech company chopped nearly half its staff and tied the cuts directly to AI capability.

This is the direction the industry is moving. Not gradually. Not as a thought experiment. As an active, profitable, board-approved strategy.

I do not think panic helps. But I think awareness does. If you are building a career in a domain that involves coordinating information, executing defined processes, or writing code in a stack where AI agents can operate autonomously, the question is not whether your company will face this pressure. The question is when, and whether you are positioned on the side that benefits from the transition or the side that absorbs it.

Defendre Solutions tracks AI industry shifts and what they mean for businesses and workers. If your organization is navigating AI-driven restructuring, get in touch. We help teams understand the landscape and build strategies that account for where the technology is actually going.