On Wednesday morning, Meta employees across the globe began receiving layoff notifications at 4 a.m. local time. By the end of the week, approximately 8,000 people will have lost their jobs — roughly 10% of the company’s total workforce. The cuts were announced last month. The waiting, according to employees who spoke to Business Insider, has been its own kind of damage. One word kept appearing in the internal descriptions of the last several weeks: dread.
But the layoff number is not the most important number in this story. The more revealing figure is 7,000 — the number of Meta employees being reassigned to AI roles before the cuts begin. In a single internal memo, Mark Zuckerberg’s company divided its workforce into two categories: the people the AI economy needs, and the people it has already decided it doesn’t.
That sorting is the story.
The framing Meta offered publicly was familiar. The cuts are necessary to help the company run “more efficiently” and offset the cost of its AI investments — investments that have been, by any measure, enormous. Meta has committed hundreds of billions of dollars to AI infrastructure, model development, and hardware buildout. Zuckerberg has made AI dominance the explicit strategic priority of the company. The layoffs, in the company’s telling, are the price of that ambition.
But efficiency is doing its usual work here. When a company generating tens of billions in annual profit cuts 10% of its workforce and calls it efficiency, the word is not describing a financial necessity. It is describing a strategic choice — a decision about what kinds of work, and what kinds of workers, the next version of the company requires.
The 7,000 reassignments make that choice visible in a way that the layoff number alone does not. Meta is not simply reducing headcount. It is actively restructuring its human architecture around AI — identifying the roles the technology enables and moving people into them, while simultaneously eliminating the roles it renders redundant. The company is not becoming an AI company. It is becoming a company whose work requires fewer humans to perform. The 8,000 cuts and the 7,000 reassignments are not separate decisions. They are two sides of the same one.
The internal atmosphere that employees described to reporters this week is worth sitting with. “Sense of dread” is not the language of a routine restructuring. It is the language of a workforce that has watched the cuts come in waves — that has seen colleagues disappear in previous rounds, survived, and now sits in the uncertainty of not knowing whether survival was temporary. Meta has now conducted significant layoffs in 2022, 2023, and 2026. Each round arrived with its own efficiency rationale. Each round left a smaller, more anxious workforce wondering when the next memo would arrive.
That psychological texture matters because it is not unique to Meta. It is the operating condition of the modern technology workforce — and increasingly of the white-collar workforce broadly. The dread is not irrational. It is a rational response to a labor environment where the terms of employment are being rewritten faster than workers can negotiate them, where the skills that made someone valuable last year are being automated this year, and where the internal sorting between AI-relevant and AI-replaceable is happening in real time, often without explanation.
The workers being reassigned to AI roles did not necessarily ask to be. The workers being cut did not necessarily deserve to be. The memo made both decisions simultaneously, in the same document, in the same week.
Meta also announced it is abandoning plans to fill 6,000 open positions — roles that were posted, that candidates applied for, that hiring managers were preparing to fill. Those 6,000 positions represent not just jobs that won’t exist, but decisions that people made — about where to apply, what to turn down, how to position themselves — based on a signal the company has now retracted.
That detail sits inside a larger pattern SSC has been tracking across the technology sector. LinkedIn cut 1,000 jobs while posting record quarterly revenue. Cisco cut nearly 4,000 people on the same day it beat Wall Street expectations. Oracle cut between 20,000 and 30,000 workers while declining to negotiate severance with the people it was letting go. The message from the sector is consistent: the numbers have never been better, and that is exactly why the cuts are possible. Growth no longer requires proportional headcount. Efficiency is no longer about survival. It is about margin.
The 7,000 reassignments are worth one more look before this story closes — because they contain an argument the coverage has largely missed.
If AI were simply replacing jobs, the story would be straightforward: technology eliminates roles, workers are displaced, the economy adjusts. That story is uncomfortable but familiar. What Meta’s memo reveals is something more specific: AI is not replacing all jobs equally. It is replacing the jobs that don’t feed the AI pipeline while creating new roles for the workers who can operate inside it. The sorting is not random. It follows the contours of existing skill distribution, educational access, and professional network — which means it follows, with uncomfortable precision, the contours of existing inequality.
The workers being moved into AI roles at Meta are not the same demographic profile as the workers being cut. The workers being cut at Meta are not the same demographic profile as the workers being cut at federal agencies. But the mechanism is identical in every case: a technology-driven restructuring that rewards proximity to the new infrastructure and penalizes distance from it.
Who gets reassigned and who gets the 4 a.m. email is not a mystery. It is a decision. And the decision has a shape.