Meta Is Hiring AI Engineers and Firing Everyone Else

Meta just cut hundreds of jobs. Its stock is up.

On Wednesday, Meta confirmed layoffs spanning several divisions — Facebook, global operations, recruiting, sales, and its virtual reality unit Reality Labs — marking the company's second significant round of cuts in 2026. Some affected employees were offered internal transfers, though in certain cases those roles come with relocation requirements. The company's statement was predictably brief: "Teams across Meta regularly restructure to ensure they're in the best position to achieve their goals."

What the statement didn't say: this is part of a deliberate pivot, and the direction it's pointing is unmistakable.

The Metaverse Is Out. Superintelligence Is In.

To understand Wednesday's cuts, you have to go back to January, when Meta laid off more than 1,000 employees in Reality Labs — roughly 10% of the division that builds Quest VR headsets and runs the Horizon Worlds virtual social platform. Multiple game studios were shut down entirely. The metaverse, which Mark Zuckerberg once bet the company's very name on, is now a shrinking line item.

The money and talent are flowing somewhere else. Meta recently inked a licensing deal with AI startup Dreamer, folding its team into the newly formed Meta Superintelligence Labs. Among those returning to the company: Hugo Barra, Dreamer's co-founder — and the same executive who led Meta's VR push from 2017 to 2021. The person who helped build the metaverse is now helping to dismantle its legacy and replace it with something new.

Meanwhile, Meta disclosed in corporate filings Tuesday that top executives — CFO Susan Li, CTO Andrew Bosworth, CPO Christopher Cox, and COO Javier Olivan — will receive a new stock option incentive program tied to aggressive 5-year performance targets. The company was unusually candid about the stakes: "This is a big bet. These pay packages will not be realized unless Meta achieves massive future success." Translation: the C-suite is being locked in for the long haul, but only if the AI gamble pays off.

Sponsored

Advertise with Us

[email protected]

Sponsored

Why the Market Cheered the Cuts

Earlier this month, Reuters reported that Meta was considering cutting more than 20% of its total workforce. Meta called it "speculative." The actual number this week is far smaller. But here's what's telling: when that Reuters report dropped, Meta's stock jumped nearly 3% in a single day. Investors didn't wait for confirmation. The mere possibility of mass layoffs was enough to send shares higher.

This is the calculus that now governs Big Tech: fewer people + more AI = better margins. It's the same logic that drove Google, Amazon, and Microsoft to cut tens of thousands of jobs over the past two years while simultaneously reporting record profits. Meta is following the same playbook, just later and louder.

The irony is sharp. The recruiting team — the people whose job is to hire — was among those let go. That's not just a cost cut. It's a signal that Meta doesn't plan to grow its headcount in the traditional sense anytime soon.

Who Wins, Who Loses

The winners here are clear: AI researchers, generative model engineers, and agent developers. Meta has been aggressively poaching talent from OpenAI, Anthropic, and Google DeepMind, and that hiring hasn't slowed. The losers are the people in operational, support, and go-to-market roles — the infrastructure of the old Meta that built and sold social media products.

For current and former Meta employees, the message is uncomfortable but legible: your value to the company is increasingly measured by your proximity to AI development. Everyone else is a candidate for the next restructuring memo.

For investors, the short-term read is positive — leaner cost structure, focused AI investment, executive retention. But the longer-term question is harder: Meta's AI ambitions are enormous, yet its path to monetizing AI beyond advertising remains murky. The stock options for top executives only pay out if the share price "meaningfully exceeds" a high exercise price within five years. That's a bet on AI delivering real revenue, not just efficiency gains.