Zuckerberg Admits Meta Got It Wrong: Here's What the AI Layoff Mess Actually Looked Like From the Inside

In May, Meta cut 8,000 jobs (around 10% of its global workforce) and moved another 7,000 employees into AI-focused roles. One in five Meta employees had their job changed or eliminated, all in the name of building the company's AI future.

Then, on 12 June, an internal memo from Mark Zuckerberg leaked. The headline quote: "Given the complexity of these changes, we've made mistakes and will almost certainly make more."

To be fair to Zuckerberg, it was a more honest admission than we usually get from Big Tech. He acknowledged that some of the new team structures weren't working, specifically, a setup in the Applied AI Engineering unit where a single manager could oversee up to 50 individual contributors. That ratio, by any measure, is chaos. Meta is now pulling back on it.

His defence of the layoffs was that by creating new AI-focused roles, Meta gave itself flexibility. If a restructured team wasn't working, they could transfer people rather than cutting them again. It sounds reasonable in theory. In practice, 8,000 people lost their jobs and 7,000 more had their roles reshuffled, all while Meta projected capital expenditure of up to $145 billion for 2026 – more than double what it spent in 2025.

Zuckerberg has promised there will be no more mass layoffs this year. To boost morale, Meta is increasing budgets for team events and running a company-wide hackathon in July.

Whether a hackathon fixes the trust problem is another question entirely.