Meta lays off 16,000 people. Its stock rises 3%

Meta has nearly 79,000 employees. It plans to get rid of about 15,800. And when Reuters published the news last Friday, the company's stock rose by nearly 3%. If anyone needs a single image to summarize what we value about human labor in 2026, there it is.

The potential job losses, equivalent to around 16,000 positions, aim to offset the costs of a planned $600 billion investment in AI infrastructure through 2028. It is not a rounded number chosen to sound ambitious: Meta already revealed in its fourth-quarter earnings report that its capital expenditure linked to AI will be between $115 billion and $135 billion in 2026 alone, roughly double what it invested in 2025.

Meta executives relayed the signal to their senior managers last week, asking them to start planning where and how to apply the cuts. Spokesperson Andy Stone responded to Reuters, describing the reports as "speculative reporting on theoretical approaches." A denial that denies nothing concrete, which is exactly the kind of refutation issued when the plans already exist but the date is not yet set.

The paradox of the moment has a code name: Avocado. Meta planned to launch Behemoth, the most powerful version of Llama 4, in the summer. It canceled it. Now its superintelligence team is working on a model called Avocado, but its performance has also fallen below internal expectations.

It's a detail that doesn't fit well with the narrative of "historic investment." The company that justifies firing 16,000 people because AI does everything better has two consecutive models that fail to meet their own goals. Despite this, Meta pushes forward: it acquired Moltbook, a social network designed for AI agents, and is in the process of buying the Chinese startup Manus for at least $2 billion. Technology Org Furthermore, last year it invested $14.3 billion in Scale AI and ended up hiring its CEO, Alexandr Wang, along with several of its top researchers.

Zuckerberg's argument is simple. In January, he said he was starting to see "projects that previously required large teams now completed by a single very talented person." Read in January, it sounds like a vision of the future. Read alongside a 20% layoff plan, that sentence operates differently.

Jefferies analysts published a note on Sunday with a direct reading: "If Meta is willing to reduce headcount on this scale while increasing investment in AI, we believe it signals a broader structural shift: AI is driving productivity." In today's financial logic, firing 16,000 people and spending $135 billion on machines is not a contradiction. It is exactly the model Wall Street wants to see replicated.

And it's not just Meta. Amazon eliminated 16,000 positions in January. Block, Jack Dorsey's fintech, made cuts in February and explicitly attributed the layoffs to AI. Atlassian announced this week the dismissal of 10% of its workforce, about 1,600 employees, redirecting that investment towards automated systems.

Consulting firm Challenger Gray & Christmas estimates that over 12,000 layoffs have been attributed to AI in the United States alone so far in 2026. The tech sector has accumulated 166 workforce reduction events this year, affecting 55,775 people as of March 14.

Almost 764 people a day. Not counting weekends.

Zuckerberg dubbed the period of cuts in 2022 and 2023 the "year of efficiency," a phrase that accompanied 11,000 layoffs in November of that year and another 10,000 four months later. The slogan aged with all the grace of NFTs. And yet, here we are again with an identical logic, this time with a new name: AI.

What has changed is the alibi. Before, cuts were justified by the need to adjust after massive expansion during the pandemic. Now, the justification is progress. Automation as a moral argument. Efficiency as an inevitable destiny.

There is something structurally ironic about the fact that the technology sold for years as a tool for the democratization of knowledge and human connection is now the argument for emptying offices. Meta built its valuation on the attention of billions of people. Now it proposes that those same people be replaced, at least in part, by the systems it is buying at the knockdown price of a national budget.

The plans have neither a confirmed date nor definitive scale. Meta still has not announced anything officially. But the sector is already acting as if it were a fact: the 16,000 potentially affected jobs would represent the largest single layoff in the tech sector in 2026, equivalent to 21 days of the industry's daily average of cuts concentrated in a single announcement.

The operative question is no longer whether Meta will cut staff. It is whether the $600 billion invested in data centers and models that are not yet yielding results will, eventually, produce something that justifies having paid for them with 16,000 professional careers. The history of major technological investment cycles suggests that some gamblers get it right. The history of those who pay the price for those wagers suggests they are rarely the same people.