Eight thousand layoff emails went out in a single wave. Another seven thousand employees got a different kind of message: pack up your old job, you have a new one on an AI team. Meta’s restructuring this year is not a routine cost-cutting exercise. It is a company physically reshaping itself around artificial intelligence, department by department.
The numbers behind the shakeup
Meta began notifying roughly 8,000 employees in May that their roles were being eliminated, a cut equal to about 10 percent of its workforce. It was the latest and largest round in a series of reductions that, combined with earlier layoffs dating back to 2022, brings Mark Zuckerberg’s total headcount cuts at the company to around 25,000 people. A further wave of cuts is reportedly planned for the second half of 2026, though the timing and scope have not been finalized.
Chief People Officer Janelle Gale told staff that more than 7,000 workers would be reassigned rather than let go, moved into newly created AI-focused teams with names like Applied AI Engineering, Agent Transformation Accelerator XFN, and Central Analytics. Combined, the layoffs and the reassignments touch close to 20 percent of Meta’s total workforce, a scale of internal reshuffling that is unusual even by the standards of a tech industry that has spent the last three years cutting headcount.
What is funding the shift
The restructuring is tied directly to money. Meta’s projected capital expenditures for 2026 range from 125 billion to 145 billion dollars, more than double what the company spent in 2025. A large share of that is going toward AI infrastructure, chips, data centers and the compute needed to train and run increasingly large models, with infrastructure costs alone estimated between 115 and 135 billion dollars for the year. That figure alone exceeds the entire annual revenue of most Fortune 500 companies, spent in a single year on servers, chips and the power to run them.
Wall Street has largely rewarded the spending so far, even as it has questioned whether the returns will materialize on the timeline Meta is promising. Investors have grown more comfortable with the idea that AI infrastructure spending is not discretionary but existential, a cost of staying competitive rather than a bet that might not pay off. That shift in sentiment gives Meta more room to keep spending aggressively, even as it cuts jobs to help fund it.
Funding that kind of buildout while keeping margins intact means finding savings elsewhere, and Meta has chosen to find them in headcount. The logic Zuckerberg has articulated publicly is that AI tools are making individual engineers and teams more productive, which means the company can do more with fewer people in some functions while it pours resources into the AI bets it considers existential to its future, from the Llama model family to AI-driven ad targeting and the long-running push into AI companions and agents.
A familiar playbook, bigger stakes
Meta is not the only company treating AI as both a cost center and a reason to cut costs elsewhere. Amazon, Google and Microsoft have all announced layoffs in overlapping periods while simultaneously ramping up AI capital spending, a pattern that has become common enough that industry analysts now describe it as standard practice rather than an anomaly. What sets Meta apart is the scale of internal reassignment alongside the cuts. Rather than simply eliminating roles and hiring fresh AI talent externally, the company is trying to retrain and redeploy a meaningful chunk of its existing workforce into AI-adjacent jobs, a bet that institutional knowledge of Meta’s products and infrastructure is worth preserving even as the skill requirements shift.
Employees who spoke to reporters described a tense spring, with entire teams uncertain until the emails arrived whether they would be cut or reassigned. Zuckerberg has said publicly that he does not expect further broad cuts on the scale of the May layoffs, even as the company acknowledges more targeted reductions are still coming later in the year.
Even Meta is fighting for compute
The restructuring is happening against a backdrop that reveals just how tight AI computing capacity has become across the industry. Google reportedly limited Meta’s access to Gemini model capacity after Meta sought more computing power than Google was able to provide, a bottleneck that disrupted and delayed some of Meta’s internal AI projects. That a company spending well over 100 billion dollars a year on its own AI infrastructure would still find itself rationed by an external cloud provider underscores how scarce high-end AI compute remains in 2026, even for the biggest players in the industry. It also helps explain why Meta is pouring so much of its own capital expenditure into building infrastructure it fully controls, rather than depending on partners for capacity it cannot guarantee.
What it signals for the rest of the industry
Meta’s overhaul is a useful data point for anyone trying to gauge how seriously large tech companies are betting on AI reshaping their own organizations, not just their products. Redirecting a fifth of your workforce toward AI roles in a single year is not a symbolic gesture. It is a statement that the company expects AI capability to be the primary determinant of its competitive position for the rest of the decade, and that everything else, including thousands of jobs, is negotiable in service of that bet.
For more coverage of how AI is reshaping the tech workforce, visit Mylistingo.
