Meta cuts 7,800 jobs while shifting 7,000 staff to AI

Big Tech is spending tens of billions of dollars on data centers and GPUs, and CFOs have to find a way to balance the ledger. Meta's latest answer is as blunt as it gets: cut older operational roles to fund the next frontier of computing.

According to an internal email obtained by Business Insider, Meta notified affected staff on Wednesday that roughly 7,800 jobs are being eliminated, representing about 10% of its 78,000-person global workforce. Meta declined to comment on the matter.

The leadership's memo didn't hide behind corporate jargon. It explicitly told staff that the head count reduction was part of an effort to run more efficiently and "offset the other investments we're making." In plain English, soaring capital expenditure on AI infrastructure has to be matched by aggressive cuts elsewhere.

Badges cut immediately, but severance stays soft

As soon as the notices landed in inboxes on Wednesday morning, affected employees saw their internal access stripped and office badges deactivated. Those already at their desks were told to pack up personal belongings and head home.

Despite the sudden lock-out, the severance package remains generous by industry standards. Impacted workers enter a paid "non-working notice period" with full base pay and continued benefit accruals, alongside regular RSU vesting through their termination dates.

The final package provides 16 weeks of base severance pay, plus an additional two weeks for every completed year of service. Meta is also covering 18 months of COBRA healthcare premiums, three months of outplacement career support via Lee Hecht Harrison, and dedicated legal assistance for visa holders. Staff can keep their jobs if they land another open internal role at least a week before their official separation, which voids the severance payout.

The great 7,000-worker pivot

This isn't a traditional contraction—it's a massive corporate reshuffle.

Alongside cutting around 7,800 workers, Meta confirmed plans to shift over 7,000 existing employees into brand-new AI initiatives. The company isn't necessarily shrinking its total footprint; it is systematically reallocating human capital toward generative AI, open-source models, and agent architectures.

Put simply, Meta no longer has room for bloated legacy support or secondary projects. If your work doesn't directly feed the AI flywheel, your headcount budget is likely being repurposed to hire machine learning researchers or pay for cluster cooling.

Why it matters

We think this move marks a brutal new phase in the enterprise AI arms race. Tech giants are no longer funding AI out of windfall discretionary cash; they are cannibalizing their own core headcount to bankroll compute.

For tech workers and developers, the implications are stark. Proximity to AI execution is now the primary metric of job safety. Building solid software in a traditional division is no longer a defense if leadership decides that budget line belongs to a cluster of clusters instead. Moving your skillset closer to model training, evaluation, and practical deployment is now essential.

For the industry, Meta's strategy puts enormous pressure on competitors like Google, Amazon, and Microsoft. By publicly showing Wall Street that it can absorb historic AI CapEx while trimming operational fat, Meta sets a precedent that shareholders will expect everyone else to follow.

From a corporate perspective, the move looks fiscally disciplined, but moving 7,000 people across teams overnight comes with severe execution risks. Retraining thousands of workers for cutting-edge AI problems is rarely seamless, and cultural fatigue could easily derail Meta's aggressive timeline.