Meta Is Reportedly Cutting 20% of Its Workforce to Pay for AI. Here Is What We Actually Know

Sources: Reuters (original report, three sources), CNBC, TechCrunch, Fox Business, LatestLY as of March 15, 2026. Meta has not confirmed these plans. All layoff details are from Reuters sourcing and are unconfirmed by Meta.

Table of Contents

• What Reuters Reported and What Meta Said Back
• The Numbers: 15,800 Jobs, 79,000 Employees, $600 Billion
• Why Is This Happening Now Specifically?
• Paying Hundreds of Millions to Hire AI Researchers While Cutting Everyone Else
• Meta Is Not Alone: The Broader Tech Layoff Wave of 2026
• The “AI-Washing” Problem: Is AI Really the Reason?
• What Zuckerberg Has Been Saying Publicly
• What This Actually Means for the People Involved
• Where This Is Going

What Reuters Reported and What Meta Said Back

Reuters published a report on March 14, 2026, citing three sources familiar with the matter, that Meta is planning layoffs that could affect 20 percent or more of its total workforce. The report says top Meta executives have already shared these plans with senior leaders inside the company. The timing and final size of the cuts have not been decided, according to the same sources.

Meta pushed back immediately. Spokesperson Andy Stone told Reuters: “This is speculative reporting about theoretical approaches.” That is a carefully worded denial. It does not say the discussions did not happen. It does not say layoffs are off the table. It says the report is speculative about approaches that are described as theoretical. That phrasing is worth reading carefully because it is doing a lot of work without actually saying no.

Whether or not 20 percent turns out to be the final number, something is clearly being planned. You do not brief senior leaders on a restructuring of this scale if it is purely hypothetical. Reuters staked the story on three separate sources inside the company. Given Reuters’ track record on breaking major corporate restructuring news and the detail level of their sourcing here, this is being taken seriously across the industry.

The Numbers: 15,800 Jobs, 79,000 Employees, $600 Billion

Meta employed 78,800 people as of December 31, 2025, according to its most recent SEC filing. Twenty percent of that is approximately 15,800 people. That would make this the single largest workforce reduction in Meta’s history, exceeding the two previous rounds combined. In November 2022 Meta cut 11,000 workers, roughly 13 percent of its workforce at the time, in a round Zuckerberg called the beginning of the company’s “year of efficiency.” A few months later it announced 10,000 more cuts. Those two rounds together were roughly 21,000 jobs but spread across two separate events.

A single 15,800-person cut would be a different kind of signal. It would represent a faster, more decisive restructuring than the drawn-out 2022 to 2023 process. And it is happening while the company is simultaneously committing to spend $600 billion on AI data center construction by 2028. That juxtaposition is the core of why this story is generating so much attention. Meta is preparing to spend more money than most countries’ annual GDP on AI infrastructure while telling nearly one in five of its employees they may no longer have jobs.

Why Is This Happening Now Specifically?

The timing comes down to a collision between two things happening simultaneously at Meta. On one side, the cost of building and running AI infrastructure has accelerated dramatically. Training frontier AI models requires clusters of tens of thousands of GPUs running for months. The data centers that house those clusters require land, power contracts, cooling systems, and specialized construction that are all expensive and take years to plan. Meta committed publicly to $65 billion in capital expenditures in 2025 alone, much of it for AI infrastructure. The 2028 figure of $600 billion is the cumulative projection of where that spending goes if the trajectory continues.

On the other side, Meta’s revenue growth, while still strong, has not kept pace with the rate at which AI infrastructure costs are rising. The company cannot simply grow its way out of the math. At some point the only way to maintain margins while spending aggressively on AI hardware is to reduce the labor cost that makes up a large portion of operating expenses. That is the logic Reuters sources are describing: the layoffs are explicitly framed as a way to offset AI infrastructure costs, not as a response to poor business performance.

There is also the efficiency argument, which Zuckerberg has been making publicly for over a year. As AI tools become capable of handling work that previously required teams of people, the company genuinely needs fewer employees to do the same amount of work. Or at least, that is the argument. Whether that is the actual driver or a convenient justification for cuts that would have happened anyway is a legitimate question that we will come back to.

Paying Hundreds of Millions to Hire AI Researchers While Cutting Everyone Else

One of the most striking details in Reuters’ reporting is the contrast between what Meta is spending to acquire AI talent and what it is reportedly planning to do to the rest of its workforce. According to Reuters, Meta has been offering pay packages worth hundreds of millions of dollars over four years to court top AI researchers to a new superintelligence team that Zuckerberg has been building.

That is not a typo. Individual AI researchers being offered packages that would make them nine-figure earners over four years, at the same company that is reportedly preparing to cut 15,800 other employees to offset costs. The gap between what Meta is willing to pay for the capabilities it considers essential for the future and what it considers expendable is unusually stark and public.

This week Meta also acquired Moltbook, a social networking platform built specifically for AI agents rather than human users. Separately, Reuters had previously reported that Meta is spending at least $2 billion to acquire Manus, a Chinese AI startup. The acquisition spending and the superintelligence hiring and the data center construction are all happening simultaneously with the reported workforce reduction planning. The picture this creates is of a company that has made a very specific bet on what the future of its business looks like and is restructuring everything around that bet as fast as it can.

Meta Is Not Alone: The Broader Tech Layoff Wave of 2026

Meta is the biggest name in a pattern that has been building across the tech industry for several months. In January 2026, Amazon cut approximately 16,000 jobs and at the time signaled that additional reductions could follow. Amazon cited efficiency gains from AI tools and broader organizational changes as the reason. Block, the payments company formerly known as Square, announced sweeping layoffs recently under similar framing. Netflix has made cuts. The list is growing week by week.

The common thread in how these companies are explaining the cuts is AI. The argument in each case is roughly the same: AI tools now handle work that previously required people, and the company needs to reflect that in its headcount. That explanation is intuitive, it fits the public narrative about AI’s impact on employment, and it is the kind of thing that analysts and investors tend to receive positively because it suggests the company is managing costs intelligently rather than having a business problem.

The scale of what is happening in 2026 is worth acknowledging plainly. If Meta’s reported 20 percent cut happens, combined with Amazon’s cuts, Block’s cuts, and the other ongoing reductions across the industry, the total number of technology jobs eliminated in a roughly twelve-month window will be in the hundreds of thousands. Many of these are not entry-level positions. They include experienced engineers, product managers, designers, and operational staff who built careers at companies that were considered among the most stable employers in the world two years ago.

The “AI-Washing” Problem: Is AI Really the Reason?

Not everyone is accepting the AI explanation at face value, and the skepticism is coming from credible places. TechCrunch noted that Sam Altman, the CEO of OpenAI, has publicly suggested that many of the layoffs happening across the tech industry may be what he called “AI-washing,” where executives use AI as a convenient and sympathetic explanation for workforce reductions that are actually driven by other factors.

The other factor that gets pointed to most often is pandemic-era overhiring. Between 2020 and 2022, tech companies hired aggressively as demand for digital services surged with remote work and COVID-era consumer behavior changes. When that demand normalized, many companies found themselves with workforces sized for a growth rate that had not materialized. The 2022 to 2023 layoff wave was widely attributed to this overhiring correction. The argument from skeptics is that what is being called AI-driven restructuring in 2026 is partly the continuation of that same correction, with AI attached to the explanation because it is a more forward-looking story than “we hired too many people in 2021.”

There is also a financial signaling dimension. Telling investors you are cutting headcount to fund AI investment reads better than telling them you are cutting headcount because revenue growth is slowing or margins are under pressure. AI-framed restructuring says “we are investing in the future.” A plain cost-cutting narrative says “we have a cost problem.” The stock market tends to react differently to these two framings even when the underlying action, reducing payroll expense by cutting jobs – is identical.

The honest answer is probably that all of these things are true simultaneously to varying degrees depending on the company. Some of the cuts are genuinely enabled by AI tools that have automated work that used to require people. Some are pandemic overhiring correction that was going to happen regardless. Some are margin management dressed up in AI language. The ratio is different at each company and probably different across different departments within the same company. Meta’s case almost certainly contains elements of all three.

What Zuckerberg Has Been Saying Publicly

The reported layoffs did not come out of nowhere. Zuckerberg has been laying the public groundwork for this kind of restructuring for over a year. In January 2025 he posted a video on Facebook and Instagram in which he said he planned to get rid of “lower performers” at Meta and replace some roles with AI. He described his goal as making Meta “the most advanced AI company in the world” and said that achieving that required both aggressive investment and a leaner, higher-performing team.

In subsequent earnings calls and public appearances, Zuckerberg has repeatedly described AI as both the company’s biggest investment priority and the thing that will make the company more efficient over time. The framing he uses consistently is that the same AI tools Meta is building and selling will also reduce the internal headcount Meta needs to operate. That argument has a certain internal logic. If your company is one of the leading developers of AI automation tools, and you are not using those tools to automate your own operations, that is a credibility problem. Zuckerberg appears to believe genuinely that the answer to rising AI infrastructure costs is partly to use AI to reduce the human operational costs the infrastructure was previously supporting.

Whether that logic actually plays out the way the CEO expects, or whether it is partially a rationalization for cuts driven by financial pressure, is something the next few quarters of earnings data will make clearer. The combination of $600 billion in committed infrastructure spending and a potential 15,800-person headcount reduction is a very large bet in a very specific direction. Zuckerberg has made large bets before. The metaverse is the obvious example of one that did not go as planned. AI is a different kind of bet in the sense that the technology is clearly working and being adopted broadly, but the financial returns at the scale Meta is committing to are not guaranteed.

What This Actually Means for the People Involved

It is worth pausing on the human dimension of what 15,800 job cuts actually means before getting too deep into the strategic analysis. These are engineers, designers, project managers, data scientists, operations staff, and people in hundreds of other roles who have in many cases spent years building their careers at Meta. The company has historically been considered one of the more desirable employers in tech, with competitive compensation, strong benefits, and work that has had genuine scale and impact. For people in that position, the news that you may be part of a 20 percent reduction is genuinely disruptive in ways that are not captured by any strategic or financial analysis.

The job market for displaced tech workers in 2026 is also meaningfully different from what it was in 2019 or 2020. Many of the companies that absorbed large numbers of laid-off tech workers in previous cycles are themselves in the middle of restructuring. The roles that are most in demand are AI-adjacent, which helps some displaced workers and not others depending on their background. The premium placed on AI expertise is creating a bifurcated labor market where the top tier of AI specialists can essentially name their compensation while workers in adjacent roles that AI is beginning to touch face genuine uncertainty about where the next opportunity comes from.

Where This Is Going

The Reuters report is from sources describing plans that have been shared with senior leaders. That is a meaningful level of specificity. It is not anonymous speculation about a company’s general direction. It is reporting on specific internal discussions that have already happened. That does not mean the 20 percent figure will be the final number, or that the timeline Reuters implies is accurate, or that the cuts will not be significantly modified before they are announced. Corporate restructuring plans change between the internal planning stage and execution. But the direction is clear enough that the “speculative reporting about theoretical approaches” denial from Meta’s spokesperson reads as damage control rather than a factual correction.

What happens next is likely a formal announcement within the next few weeks to months, followed by a more specific breakdown of which teams and roles are affected. The framing of that announcement will matter, both for how the affected employees receive the news and for how the market interprets the strategic rationale. If Meta frames this as a forward-looking AI transformation, the stock may respond positively despite the scale of the cuts, much the way Amazon’s stock has held up through its own reduction announcements. If the financial pressure context becomes more prominent than the AI transformation story, the reaction will be different.

For anyone working in tech right now, regardless of whether they work at Meta, this is the story that makes the “AI is going to change jobs” conversation feel concrete and immediate rather than theoretical. The question of whether AI is taking jobs or whether companies are using AI as a story to explain cuts that would have happened anyway matters less to the people holding a layoff notice than the fact of the notice itself. That is the part of this worth not losing sight of in the coverage.

What do you think is actually driving this: genuine AI efficiency, or overhiring correction dressed up in AI language, or both? Drop your take in the comments. This is one of those situations where the official narrative and the real explanation are probably pretty different things.

References (March 15, 2026):
Reuters original report (Katie Paul, Jeff Horwitz, Deepa Seetharaman), March 14, 2026: reuters.com
CNBC coverage of Reuters report: cnbc.com
TechCrunch coverage including Sam Altman “AI-washing” quote: techcrunch.com
Fox Business on Meta spokesperson statement and workforce figures: foxbusiness.com
Meta 2025 annual filing (79,000 employee count as of December 31, 2025): SEC EDGAR
Reuters on Meta’s $2 billion Manus acquisition: reuters.com (previously reported)
Meta $600 billion data center commitment by 2028: Meta investor communications

Meta is spending $600 billion to build the AI future.
Roughly 15,800 of its employees may not be part of it.