Meta is moving thousands of employees into newly formed artificial intelligence groups as the company prepares another round of layoffs, according to The Information. The reorganization, reported by The Information, marks one of the largest internal reshuffles at the company since Mark Zuckerberg launched his aggressive push to chase frontier AI capabilities earlier this year.
The reshuffle ties directly into Meta’s Superintelligence Labs effort, the umbrella organization Zuckerberg built to consolidate the company’s scattered AI work. Workers from across the org are being pulled into new teams, while others are expected to be cut as Meta trims headcount in areas it no longer considers strategic.
What’s happening
Here’s the shape of the move based on The Information’s reporting:
- Scale: Thousands of employees are being reassigned, not just a handful of senior researchers.
- Direction: Staff are flowing into new AI-focused groups tied to Meta’s superintelligence push.
- Cost: Layoffs are coming alongside the shuffle, signaling this isn’t pure expansion. It’s a rebuild.
The pattern matches what Meta did across its Reality Labs and infrastructure divisions earlier this cycle. Zuckerberg is concentrating talent and budget on a narrower set of bets, and people whose work doesn’t fit the new structure are at risk.
Why this matters
Meta has spent the last 18 months trying to catch OpenAI, Anthropic, and Google in frontier model development. The company poured billions into compute, paid eye-watering packages to recruit top researchers from rivals, and folded its FAIR research lab into the Superintelligence Labs structure. The reshuffle reported by The Information is the operational follow-through on those bets.
What stands out here is the layoff angle. Meta isn’t just hiring superstars and adding teams on top of what already exists. It’s clearing space. That tells you Zuckerberg is treating AI talent allocation as a zero-sum problem inside the company: every researcher and engineer either fits the new AI roadmap or doesn’t.
This is also the second major AI reorganization at Meta in less than a year. The previous restructuring centralized model training under Alexandr Wang after Meta’s $14 billion Scale AI deal. The current move appears to push that consolidation deeper into product and applied teams.
Context: how Meta got here
A quick recap of the status quo before this shift:
- Llama stumbles. Llama 4 launched to a cool reception earlier in the year, and Meta’s open-weights strategy looked less commanding than it did during the Llama 2 and 3 era.
- Superintelligence pivot. Zuckerberg responded by spinning up Superintelligence Labs, recruiting Wang and others, and putting frontier model work under tighter executive control.
- Spending escalation. Meta guided capex above $100 billion for 2026, most of it AI infrastructure.
- Now: people. With compute and leadership in place, the company is moving the workforce to match.
The sequencing is logical. You can’t run a frontier lab if your org chart still reflects last year’s priorities.
What to watch next
For practitioners and operators tracking this, a few things to expect:
- More poaching pressure on rivals. Meta’s new teams need senior ML and infra talent. Expect another wave of competitive offers aimed at OpenAI, Anthropic, and Google DeepMind staff.
- Faster product shipping. Consolidated AI teams usually mean shorter paths from research to product. Watch for new consumer AI features in Instagram, WhatsApp, and Ray-Ban Meta smart glasses.
- Open-source signal. Whether Meta keeps releasing open-weights models, or quietly steps back, will be a major tell about how the new structure is positioned versus closed-model competitors.
- Morale risk. Layoffs paired with reorgs tend to push out mid-career talent first. Meta’s AI team has already seen high-profile departures this year.
The broader takeaway is that the big labs are converging on the same playbook: fewer, larger, better-funded teams pointed at a single frontier goal. Meta is now executing on that pattern at the workforce level.
Full details at The Information.