Mark Zuckerberg is now admitting Meta’s AI overhaul came with mistakes, including strained team structures, mass role changes and a workforce reset that left employees bracing for what came next. The Meta AI layoffs hit in May, when the company cut 10% of its global workforce and moved 7,000 employees into new AI-related initiatives, according to Reuters.
In an internal memo reviewed by Reuters, Zuckerberg told employees, “Given the complexity of these changes, we’ve made mistakes and will almost certainly make more.” He added that he is “focused on providing as much stability as possible” while the company reworks itself around artificial intelligence.
That stability has been the hard part. Meta’s May restructuring affected roughly 20% of the company through layoffs and transfers. The company also closed 6,000 open roles, while employees pushed back internally over the changes, including concerns tied to mouse-tracking software being used to train AI systems.
Zuckerberg also tried to cool fears about another sweeping cut.
“I don’t want to overpromise because the world is changing in ways that are out of our control,” he said, while repeating that Meta does not expect more company-wide layoffs this year.
The issue is not just who got cut. It is who got moved, where they landed and how fast Meta tried to flatten parts of the company while racing deeper into AI. Meta’s new Applied AI Engineering unit had a structure where managers could oversee up to 50 individual contributors. Zuckerberg now says Meta plans to pull back on that approach.
Meta is also trying to rebuild some internal trust with bigger budgets for offsites and corporate events, plus a July hackathon meant to push collaboration around its latest models. Still, the numbers show why the pressure is not easing. Meta raised its 2026 capital spending forecast to $125 billion to $145 billion, citing higher component pricing and added data center costs for future capacity.
Zuckerberg is not walking back the AI gamble. He is admitting the rollout got rough while doubling down on the machine. Meta is spending like the future depends on AI, but thousands of workers already paid the first bill.
