Sometime before the end of this year, tens of thousands of people at Meta are expected to receive the kind of message that has become distressingly familiar in tech: your role is being eliminated. The company that spent years hiring aggressively, then cut 21,000 jobs in 2022 and 2023 to prove it had learned its lesson, is apparently preparing to do it again. Only bigger.
Reuters reported on March 14 that Meta's top executives have told senior leaders to start planning how to "pare back" the workforce by more than 20%. At Meta's current headcount of 78,865 employees, that number works out to roughly 15,800 to 16,000 people. No final date has been set. No exact figure has been confirmed. But the direction is clear enough that it moved markets.
Wall Street Rewarded the Plan Before It Happened
Meta's stock fell 3.8% on Friday as investors processed the initial Reuters report with some uncertainty. Then it climbed nearly 3% in premarket trading on Sunday night and Monday morning. By 5:20 AM EDT on March 16, shares were trading at $633.01, up $19.83 from Friday's close.
The reaction captures something important about how public markets now think about layoffs at large technology companies. Cutting 16,000 people is not viewed as a sign of distress. It is viewed as financial discipline. Investors read it as a signal that Meta is choosing to fund its AI bets by taking costs out of the business rather than by issuing debt or diluting shareholders.
Jefferies analysts put it plainly: "If Meta is willing to reduce headcount at this scale while ramping AI investment, we think it signals a broader shift — AI is increasingly driving productivity." Guggenheim carries a Buy rating on the stock.
The Numbers Behind the Decision
Meta's 2026 capital expenditure guidance, issued in January during its Q4 2025 earnings call, shocked analysts: $115 billion to $135 billion. That is nearly double the $72.2 billion Meta spent on capex in 2025, which was itself up nearly 70% from 2024.
The long-range figure is even more striking. Zuckerberg has committed Meta to spending at least $600 billion on US data centers and related AI infrastructure through 2028. That includes a new 2,250-acre site in Louisiana called Hyperion, estimated to cost $10 billion and deliver 5 gigawatts of compute capacity. It includes the deployment of more than one million GPUs to support next-generation AI models. It includes Meta Superintelligence Labs, the division Zuckerberg founded in July 2025 and has been personally staffing by offering nine-figure compensation packages to researchers he meets at his homes in Palo Alto and Lake Tahoe.
Full-year 2026 total expenses are expected to land between $162 and $169 billion. To keep operating income above 2025 levels despite that spending, something has to give. The headcount is the something.
What Zuckerberg Has Been Saying
Zuckerberg did not announce the layoffs himself. He did not need to. He telegraphed them in January.
On the Q4 2025 earnings call, he told investors he was beginning to see "projects that used to require big teams now be accomplished by a single very talented person." That sentence, delivered quietly and without emphasis, is about as clear a statement of intent as a CEO is permitted to make before the cuts are official.
He has also described 2026 as a year defined by building "personal superintelligence" — his term for AI systems that can handle complex reasoning, write code, manage tasks, and act on behalf of users. The infrastructure to support that vision costs hundreds of billions. The question was always who would pay for it.
The answer, it appears, is the employees.
How This Compares to the Broader Industry
Meta is not alone. The pattern of cutting people while announcing AI spending is playing out across the entire sector.
| Company | Layoffs in 2025–2026 | AI Capex Commitment |
|---|---|---|
| Meta | ~21,000 (2022–2023) + planned ~16,000 | $600B through 2028 |
| Amazon | 14,000 (Oct 2025) + 16,000 (Jan 2026) | Undisclosed; AWS AI expansion ongoing |
| Microsoft | Rolling reductions since 2024 | $80B in fiscal year 2025 alone |
| Voluntary exit programs across divisions | $91.4B capex in 2025 | |
| Atlassian | 1,600 (March 11, 2026) | Internal AI tooling investment |
According to data from Challenger, Gray & Christmas, AI has been explicitly cited in more than 12,000 US job cuts so far in 2026. In March alone, tech sector layoffs have hit 45,000 people, with over 9,200 of those directly attributed to AI and automation. The pattern is consistent: companies announce massive AI spending and, within months, announce headcount reductions.
The Critics Have a Point
There is a legitimate argument that these cuts are being dressed up as something they are not. AI is not yet performing the full-time work of 16,000 employees. The technology has improved faster than almost anyone predicted, but the productivity gains at scale are uneven. Writing code, summarizing documents, and drafting marketing copy are easier to automate than building products, managing teams, or maintaining relationships.
A number of analysts and labor researchers have noted that tech companies may be using AI as cover for reversing pandemic-era overhiring. "We suspect some firms are trying to dress up layoffs as a good news story rather than a bad one," one analyst observed, pointing to how companies frame cuts as AI-driven transformation rather than a correction of past excess.
The Harvard Business Review published research in January 2026 noting that companies are laying off workers "because of AI's potential — not its performance." The paper found that job losses are real and happening now even though most generative AI use cases have not yet delivered on their projected productivity gains.
Meta itself is a study in this tension. It is simultaneously cutting 16,000 people and paying nine-figure bonuses to attract the AI researchers who are supposed to make those cuts worthwhile. It is eliminating roles at one end of the org chart and creating new ones at the other. The net impact on human employment at Meta, short term, is sharply negative.
Andy Stone, a Meta spokesperson, offered the company's only public comment on the Reuters report: "This is speculative reporting about theoretical approaches."
That statement was issued on March 14. By March 16, the stock was up 3%.
The People on the Other Side of the Trade
Lost in the market reaction is a simpler set of facts. Meta employs 78,865 people. If the reports are accurate, roughly one in five of them could be told their job no longer exists. Many of those people joined Meta after the 2022–2023 cuts, believing the worst was behind them. Some are engineers. Some are product managers, designers, researchers, operations staff. Some built the features that billions of people use every day.
The 2022 cuts followed a period of overexpansion during the pandemic. This time, the rationale is different: the company does not need as many people because AI can do more of the work. That is a harder argument to push back against, and a harder situation for workers to navigate. You cannot fix an AI productivity curve by working harder.
The World Socialist Web Site noted in March 2026 that tech CEOs are "boasting about AI-driven mass layoffs," pointing to how the framing has shifted. Executives no longer apologize for cutting people. They present it as evidence of good management.
For a look at how AI tools are already changing the math on developer productivity — and not always in the ways companies claim — this piece on Meta's own research on AI-assisted developers is worth reading. And the Atlassian story from last week shows how quickly the AI-investment-as-justification framing has become standard.
The Bottom Line
Meta is building something enormous. The $600 billion commitment to AI infrastructure is real. The Hyperion data center in Louisiana is being constructed. Meta Superintelligence Labs has poached researchers from OpenAI, Google DeepMind, and Scale AI. The Manus acquisition brought in a proven AI agent platform. None of this is theater.
But the human cost of funding that build is also real. If the Reuters report proves accurate, roughly 16,000 people will lose their jobs at a company that earned $59.9 billion in revenue last quarter and whose CEO is offering billion-dollar signing bonuses to individual AI researchers.
The stock's 3% climb captures the trade investors are being asked to make: accept the short-term pain of displacement in exchange for a long-term bet on AI-driven productivity and profit. Markets said yes. The employees who will be affected did not get a vote.
The "Year of Efficiency" had a nice ring to it in 2023. What Zuckerberg is planning now does not have a name yet. It is bigger than that.
Sources
- Reuters via TradingView: Meta Planning Sweeping Layoffs as AI Costs Mount
- CNBC: Meta Stock Climbs Nearly 3% on Report of Planned Layoffs
- CNBC: Meta Planning Sweeping Layoffs as AI Costs Mount
- TechCrunch: Meta Reportedly Considering Layoffs That Could Affect 20% of the Company
- Engadget: Meta Is Reportedly Planning to Cut Up to 20 Percent of Its Staff
- Meta Investor Relations: Q4 and Full Year 2025 Results
- Seeking Alpha: Meta Confirms $600B Capex to Build AI Infrastructure in US
- Data Center Dynamics: Meta Estimates 2026 Capex Between $115–135BN
- Jefferies analyst quote via CNBC
- SiliconAngle: Report — Meta Could Lay Off 20% of Its Staff and Replace Many with AI Workers
- TechCrunch: Meta Just Bought Manus
- Amazon Layoffs January 2026 via CNN
- Tech Layoffs 2026: 45,000 in March via OpenTools
- Harvard Business Review: Companies Are Laying Off Workers Because of AI's Potential, Not Its Performance