Meta Platforms announced plans to lay off 10% of its workforce, about 8,000 people, in May to boost efficiency and offset AI investments.
The social-media company has been unusually aggressive, even by the standards of Silicon Valley, at pushing to incorporate AI into its employees' workflows and using it to streamline and accelerate its operations.
Already this year, it has started grading employees in performance reviews on their AI use; created ultraflat teams with almost no managers; and begun to develop a so-called CEO agent to assist Chief Executive Officer Mark Zuckerberg in performing his job. Zuckerberg and his lieutenants have repeatedly emphasized the way AI lets small teams do the work of large ones, while moving faster.
This week, the transformation kicked into high gear.
On Thursday, the company said it planned to lay off 10% of its workforce, or about 8,000 people, on May 20 as it seeks to achieve greater efficiency and find cost savings to balance out its huge investments in AI.
Earlier in the week, an internal memo notified employees of a new software tool that would record their keystrokes, mouse movements and click locations to teach "the next generation of our AI models to use computers."
In a separate memo, Meta's technology chief, Andrew Bosworth, said the company is building toward a vision where AI agents primarily do the work. "Our role is to direct, review and help them improve," he said.
It is all part of the tech giant's plans to become "AI native" and transform the way its teams and employees do their jobs as it seeks to spend up to $135 billion on AI infrastructure this year and build what it calls personal superintelligence for its 3.5 billion daily users. The moves have left some staff filled with anxiety and wondering: Am I helping automate my own job away?
Meta's downsizing follows announcements of layoffs at other tech companies including Amazon.com, Snap, Block and Oracle. The companies have offered differing rationales for the reductions. Block was the only one to link the cuts directly to AI: "We're already seeing that the intelligence tools we're creating and using, paired with smaller and flatter teams, are enabling a new way of working which fundamentally changes what it means to build and run a company," CEO Jack Dorsey wrote in a post on X.
Dorsey said on a call with analysts: "Within the next year, I believe the majority of companies will reach the same conclusion and make similar structural changes." Many observers have said that the costs of developing AI to date have been more directly responsible for tech-industry layoffs than the productivity gains.
The internal memo on Meta's new tracking software, which was earlier reported by Reuters, went viral on social media. Several employees posted questions and complaints on an internal Meta discussion board.
"This makes me super uncomfortable. How do we opt out?" reads the top comment on the post. (There isn't a way for employees to opt out.)
Another comment asked if employees' personal email accounts would be exempt. The answer was no.
On Blind, a website where verified employees can anonymously post about and review their employers, a user compared the tracking to so-called bossware that some companies use to monitor worker productivity. "What next? Implement chips in our brains to read our minds? How is this totally unrelated to the upcoming layoffs?" the user wrote.
Meta said the new tracking tool is meant to help Meta Superintelligence Labs, which recently released a new AI model, teach its models basic computer skills such as choosing from dropdown menus and keyboard shortcuts. A spokesman for the company said that the data won't be used for any other purpose and that there are safeguards in place for sensitive content. It won't be used in performance reviews, and managers can't access it, he added.
According to Blind's analysis of posts by Meta workers, sentiment among the company's staffers is at its most negative level on record.
In 2024, roughly 20% of the posts about Meta on Blind were negative. This year, that number has grown to more than 80%.
Janelle Gale, Meta's chief people officer, said in a memo Thursday that announcing layoffs a month ahead of notifying affected employees was "incredibly unsettling" but that it was necessary to address leaked information. "I know this is unwelcome news and confirming this puts everyone in an uneasy state, but we feel this is the best path forward, given the circumstances," she wrote.
Zuckerberg has been enthusiastic about AI's potential, spending billions of dollars last year to hire top researchers and touting the technology's role in driving growth of its advertising business. In recent months, he has increasingly focused on the way the technology can supercharge work inside the company and alter the structure of its workforce.
"We're starting to see projects that used to require big teams now be accomplished by a single very talented person,"
he said during Meta's earnings call in January.
By the time he said that, Meta had already laid off 1,500 employees in its Reality Labs division. In March, Meta created a new applied AI engineering organization with an ultraflat structure of 50 employees to one manager, and appointed Bosworth to lead the company's AI For Work initiative to get employees to use more AI. Zuckerberg has been spending more of his own time contributing to the company's code base and has been involved in developing the "CEO agent" that can retrieve information he needs from other parts of the organization,The Wall Street Journalpreviously reported.
In an essay published inside Meta last month, Bosworth said 2025 "feels like 100 years ago" as new ways of AI-enabled collaboration emerge.
"Lately I often feel like I'm working at two different companies depending on who I'm talking to,"
he wrote. "Many teams are still working the way we did in 2025: large groups producing carefully curated documents, project plans, and formal reviews.
"But a small and growing number of teams are working in a completely different way in 2026. These teams are tiny. They move extremely quickly. They communicate mostly through working demos instead of documents."
One big difference between the two types of teams, he said, is the way AI-forward teams embrace uncertainty: "Trying something, learning quickly, and iterating becomes far less risky than trying to plan everything up front."
The internal memo on the tracking software said the company will explore more ways to use employee workflows and data to build better AI in the future.