Meta began cutting 1,400 positions across Washington state on July 22, the latest in a series of reductions that have already eliminated roughly 8,000 jobs -- about 10% of its 80,000-person workforce -- earlier this year. The new round targets integrity, cybersecurity and Reality Labs teams specifically, while protecting AI infrastructure and monetization roles, a clear statement of where the company believes its future headcount needs to concentrate.
What makes the timing notable is that Meta simultaneously shifted 7,000 workers into AI-focused roles in July -- the company is shrinking overall headcount and reallocating thousands of people toward its AI effort in the same month, rather than treating AI hiring as simply additive. CEO Mark Zuckerberg has been unusually candid about why: he's acknowledged the AI push "hasn't really accelerated in the way we expected," with internal tools and coding agents developing more slowly than leadership projected when the superintelligence lab was stood up.
Meta isn't alone. Amazon has also cut jobs inside its own AGI unit this week, and the pattern across Big Tech in 2026 has been consistent: massive AI capex commitments paired with real headcount reductions in adjacent and even AI-adjacent teams, as companies try to fund infrastructure spending without matching revenue growth from the AI products themselves yet. It's a different shape than the 2023-era efficiency layoffs -- this time the cuts are explicitly framed as reallocation toward AI rather than pure cost-cutting, even when the practical effect for laid-off employees is identical.
Zuckerberg has reaffirmed superintelligence remains one of Meta's highest priorities despite the slower-than-expected internal progress, which puts the company in the position of doubling down on AI ambition publicly while trimming the org chart around it. For a company spending tens of billions on AI infrastructure and top-tier research talent, a 1,400-person cut in Washington state is a rounding error financially but a meaningful signal about internal confidence in near-term AI-driven productivity gains.
For operators and founders building AI tools sold into enterprises, the read is important: if Meta's own internal AI agents and tools haven't accelerated engineering output the way leadership expected, that's a useful data point on how far agentic coding and internal tooling actually are from delivering promised productivity gains, even inside a company with essentially unlimited compute and top research talent.
Watch whether more West Coast tech giants follow with similar "AI reallocation" framed layoffs through the rest of Q3, and whether Meta's Q3 earnings call gives investors a clearer read on whether the superintelligence lab's slower progress is a temporary bottleneck or a signal the return on Meta's AI capex is taking longer to materialize than the market has priced in.
