Meta to Cut 8,000 Jobs Wednesday in Largest AI Restructuring Yet

On Wednesday, May 20, Meta will execute the first wave of what could become the largest AI‑driven workforce restructuring in tech history. An internal document obtained by Reuters confirms that the company will lay off approximately 8,000 employees — 10% of its global workforce — in three batches across time zones starting at 4 a.m. local time.

But the layoffs are only one piece of a much larger restructuring. Meta is simultaneously closing 6,000 open roles and has already reassigned 7,000 workers into AI‑focused positions, according to internal documents reviewed by The New York Times. When you add it up, roughly 21,000 positions are affected — either eliminated, frozen, or radically redefined.

The cuts will hit teams across Reality Labs, the Facebook social division, recruiting, sales, and global operations, The Next Web reports. This is not a targeted trim — it is a companywide reorganization built around a single thesis: AI can do this work now.

Meta first announced these layoffs in an April 23 memo from Chief People Officer Janelle Gale — giving employees nearly a full month of uncertainty before the axe falls. "I know this leaves everyone with nearly a month of ambiguity, which is incredibly unsettling," Gale wrote in the memo, obtained by.³

The mood inside Meta, according to multiple employees who spoke to Business Insider, is a strange mix of resignation and dark humor. One employee described the atmosphere as "just waiting around" with no clarity on how layoff lists are being compiled. Others are so close to securing new roles that they're actively hoping to be laid off to collect severance. That same employee told:³ "It's a bit surreal that 1 out of 10 people are about to be hit, and no one knows how the lists are being made. It feels like people are just in a holding pattern waiting for Wednesday."

Former Meta employee Adel Wu captured the split sentiment on X, telling Business Insider: "My friends still there are either just waiting hoping to get laid off or extremely anxious because the job is their lifeline."

The severance package — 16 weeks of base pay plus two additional weeks per year of employment, with 18 months of COBRA healthcare coverage — is generous by industry standards, The Economic Times reports. But for employees in this job market, severance is a cold comfort — there are 113,000 other laid‑off tech workers competing for the same roles.

While 8,000 people lose their jobs Wednesday, another 7,000 have already been moved into what Meta is calling its Applied AI division. A leaked internal memo obtained by Business Insider details a sweeping rebranding of traditional engineering and product roles.

Three new job titles have replaced the old hierarchy: "AI Builder," "AI Pod Lead," and "AI Org Lead." Roughly 1,000 employees have already been rebranded, and the company is phasing out traditional middle management roles in favor of AI‑native "pods," according to an April 14 memo from Maher Saba, who heads the new Applied AI Engineering division under CTO Andrew Bosworth.

The restructuring is being led by two executives who represent Meta's new priorities. Alexandr Wang, the 28‑year‑old CEO of Scale AI whom Zuckerberg hired as Chief AI Officer in June 2025, now runs Meta Superintelligence Labs. Meta acquired a 49% stake in Scale AI for approximately $14.3 billion to secure him. Maher Saba runs Applied AI Engineering.

On an April 29 earnings call, Zuckerberg told analysts that small teams using AI tools can now build and launch projects within a week that previously needed dozens of engineers and months of work. "We're streamlining our teams so they aren't larger than necessary," Zuckerberg said, as reported by The Economic Times.

Here is the tension that has not gone unnoticed inside Meta: the company is spending between $115 billion and $135 billion on capital expenditures in 2026 — nearly double the $72.2 billion spent in 2025 — while simultaneously showing 8,000 workers the door.

That capex figure, reported by The Next Web, goes toward data centers, GPUs, and Llama infrastructure, including a $27 billion joint venture with Nebius for a gigawatt‑scale AI data center in Louisiana. CFO Susan Li warned of "significant acceleration in infrastructure expense growth" from depreciation and operating costs of all this hardware.

The contrast with executive compensation has sharpened the sting. Meta granted senior executives stock options worth up to $921 million each, tied to a $9 trillion market capitalization target by March 2031, in the same period it was laying off hundreds of employees, according to The Next Web. Bank of America projects $7–8 billion in annualized savings from the restructuring.

Zuckerberg has been explicit that these layoffs are not about survival. Meta posted $201 billion in revenue in 2025 with $60 billion in profit. The layoffs are about funding the bet — maintaining Wall Street's expected margins while pouring money into AI infrastructure.

Alongside the layoffs, Meta has implemented a new performance rating system that categorizes employees into four tiers: the top 20%, middle 70%, lower 7%, and bottom 3%. Top performers can receive up to 300% of their base bonus. But managers have been asked to mark 15 to 20% of employees as "below expectations," according to The Next Web.

Meanwhile, new internal software that tracks keystrokes and mouse movements to train AI models has caused unease among remaining staff. Employees have distributed protest flyers in offices and bathrooms, Business Insider reports. Large empty boxes have appeared in Menlo Park offices — employees speculate they are for packing belongings after the 20th.

On an April 29 earnings call, Zuckerberg attempted to distance himself from the AI‑replaces‑jobs narrative. The Economic Times reported that Zuckerberg told analysts "AI isn't going to replace people," arguing the technology will help individuals become more productive. Per,³ one employee said: "This won't be the last time it happens, so there's also this feeling that as much as it sucks, it's the new normal."

Meta's restructuring is not happening in isolation. As of May 19, 2026, 179 layoff events have impacted 113,863 tech workers this year — an average of 819 job losses per day, according to the SkillSyncer layoffs tracker. The first quarter of 2026 alone saw 81,700 layoffs, the highest quarterly figure since early 2023.

The pattern is unmistakable. Amazon cut 30,000 corporate employees (~10% of white‑collar staff), citing AI efficiency gains, Reuters reported. Block (Square, Cash App) laid off nearly half its staff — 4,000 people — with CEO Jack Dorsey explicitly linking the cuts to what he called "intelligence tools," per the.⁴ Atlassian's CEO said AI "changes the mix of skills we need." The language varies but the throughline is the same: AI isn't just augmenting work anymore — it's replacing the workers.

² captured the dynamic in its analysis: "The layoffs are not about survival. They are about funding the most expensive corporate bet in technology history while maintaining the operating margins that Wall Street expects." For builders watching this unfold, the signal is clear: the companies writing the checks are betting that AI‑driven productivity gains will more than compensate for human headcount reduction — and they're willing to bet billions on that thesis.

The Meta restructuring offers a real‑time case study in which roles are being automated and which are being created. The data is unambiguous: traditional software engineering roles are being eliminated while "AI builder" roles are expanding. The 7,000 employees being reassigned aren't being retrained as generalists — they're being funneled into a highly specific AI engineering pipeline.

Zuckerberg's own framing from the earnings call is instructive: small teams using AI tools can now build in a week what previously took months. The implication is that raw engineering headcount is no longer the bottleneck — AI tool proficiency is. Meta has reportedly set an internal target of 75% of code being written by AI tools, according to industry reports.

For builders outside Meta, the takeaway is stark: the skills that survive this pivot are AI system design, prompt engineering at scale, evaluation and testing of AI outputs, and integration architecture — not the routine coding tasks that tools like Claude Code, Cursor, and Copilot are increasingly handling autonomously. The May 20 layoffs are not just Meta's problem. They are a preview of what every tech company's workforce will look like within the next 18 months.