原文 · 未翻译
As Meta employees brace for layoffs next Wednesday, May 20, many say the vibes are horrifically, historically low. “Everyone is unhappy; the only people who are not unhappy are, literally, executives,” says an employee who works on Instagram.
The social media giant plans to cut about 10 percent of its workforce, or nearly 8,000 people, “to run the company more efficiently” and “offset the other investments” it’s making, according to a human resources leader. But the layoffs, which will add to the roughly 25,000 cuts Meta has announced over the past four years, are far from the only cause of rock-bottom morale.
Widening pay gaps among employees, courtroom losses for the company, and mandatory role changes for hundreds of top engineers have also contributed to what employees view as a uniquely grim atmosphere inside Meta. Yet another issue has been the recent installation of corporate software on employees’ computers to track their activity solely in the name of training AI, according to 16 current and former employees from a variety of roles who spoke with WIRED. They declined to be named because of company policies barring unsanctioned conversations with journalists.
“I don’t know anyone having a good time,” says a policy staffer. “The vibe is a bit ‘over it’—lack of connection to the mission, upcoming layoffs, American employees being used to train the AI models that will replace them.”
Anyone who can afford to leave is hoping to be laid off and receive the 16 weeks minimum of severance and 18 months of paid health care that come with it, several people say. As the Instagram employee put it, “Everyone is just like, do it now, jesus fucking christ.” Only the individuals with the best pay packages and involved in the core development of AI seem to be thriving, a longtime senior leader at Meta says.
In the UK, some workers have become so frustrated that they’re registering signatures to form a labor union. “Our leadership are escalating their cruel and shortsighted behaviors,” organizers inside the company wrote in a pitch to colleagues. “We need to create an incentive for them to treat us with basic humanity.”
United Tech & Allied Workers, which describes itself as the UK's largest union for tech workers, said last week that Meta employees wanted to organize with the group to protect their jobs, benefits, and privacy. Earlier this month, UK employees at Google DeepMind voted to unionize with the labor group’s parent organization, Communication Workers Union, over concerns about selling AI to the US military.
Pockets of employee protest have become a constant and defining feature of the biggest tech companies, including Meta, Amazon, and Google. But the latest concerns at Meta appear more widespread—so much so that they are apparently hindering the company's recruiting efforts, an employee alleges (Meta rejects the assertion). “There’s a lot of anger and fear,” a legal staffer adds. “It’s frustrating to watch because it feels so unnecessary”—particularly given how well Meta’s ad business continues to perform.
Meta largely declined to comment on specifics for this story but pointed to previous public statements defending its job cuts and new AI-related projects, including the tracking software. “There are safeguards in place to protect sensitive content, and the data is not used for any other purpose,” Meta spokesperson Tracy Clayton says.
Mounting Complaints
Some of the employee complaints come down to money. In February, for the second consecutive year, Meta cut the portion of annual raises that are paid in the form of company shares, trimming them by 5 percent on top of last year’s 10 percent snip. Median total compensation at Meta fell to $388,200 last year from $417,400 in 2024, according to public filings, though Meta's Clayton said salaries are still trending higher than they were in 2022. Pay has been further diminished by Meta shares falling about 5 percent this year as the company shifts focus from struggling virtual reality projects to even more costly AI development. “For many employees, salary is half stock, so that sucks,” the Instagram employee says.
The cuts to compensation and jobs have come amid back-to-back quarters of strong profits for Meta, to the tune of nearly $27 billion in the first three months of this year. And last year, Meta CEO Mark Zuckerberg offered to pay several top AI researchers as much as $100 million a year, or what a former executive calls “insane amounts of money relative to what anybody in that company has ever been making.”
The company’s total expenses in this year’s first quarter rose 35 percent from a year ago, to $33.4 billion, in part because of an apparent commitment to spend whatever it takes to secure top talent and AI compute power. (Zuckerberg just upped the company’s forecast for total capital expenditures—largely spending on data centers—by $10 billion to somewhere between $125 billion and $145 billion this year.) At the same time, employees feel that everyone is seemingly being subjected to ruthless scrutiny and internal performance-tracking.
Other recent events have contributed to sinking morale in less quantitative ways. In March, separate state court verdicts in California and New Mexico validated long-simmering ethical concerns in some parts of Meta, workers say. Jurors held Meta liable for product and policy failures that contributed to unhealthy experiences for its users and ordered it to pay a combined nearly $380 million in damages and civil penalties (Meta filed a motion in California requesting the verdict be overturned and is appealing the verdict in New Mexico). For some employees, the trials were an uncomfortable reminder of the harm Meta services have caused.
Early last month, Meta began forcibly moving at least 1,000 top engineers to a new Applied AI Engineering division working on tools and data to help research scientists develop better generative AI models. Anyone refusing a transfer faced the prospect of a layoff, an unusual threat in Silicon Valley where technical employees typically have the option to move to other teams during a restructuring. Some people at Meta called it a “draft,” an employee says. The incident made it evident to one technical employee that Meta was no longer “seeing us as partners.”
Purgatory
News of the impending 10 percent cuts leaked in March of this year. But the company didn’t confirm the layoffs for weeks, several employees say. Zuckerberg finally addressed the layoffs in a companywide meeting last month. The message, as some employees understood it, was that he would prefer to keep everyone, but given that AI costs so much to develop, his hands were tied. Employees would find out in May whether they still had a job. In the interim, many were left to spiral.
About the same time, the company introduced mandatory software onto corporate laptops that tracks what US employees are typing and clicking to gather data to train AI models that execute tasks such as web browsing or organizing folders on a computer as a human would. Opting out is not possible, according to three employees. “Nobody is happy about it,” says a current employee. “And we have no choice.” Some employees claim they have found work-arounds to dodge tracking or have managed to delay installation.
The software, known as Model Capability Initiative, or MCI, suddenly turned people across the company into privacy zealots, a legal staffer says. When employees protested the rollout in internal messages, including by referencing Meta’s history of user data breaches, chief technology officer Andrew Bosworth “belittled and berated” the dissenters, one veteran employee says and another confirms. “These billionaires can’t even feign empathy,” the first person says. “The social contract is completely shattered at this point.”
Anger surged when, sources say, Zuckerberg seemed to imply during an internal discussion that Meta preferred to monitor staff over external contractors because its employees were smarter.
Employees are still fighting back. This week, a small protest group posted flyers throughout several US offices urging colleagues to sign a petition that calls for Meta to end use of the tracking program. They cited “serious concerns around privacy, consent, and trust in the workplace” and contended that “any approach to AI that relies on intrusive, coercive, nonconsensual data collection contradicts” Meta’s responsible AI principles. Reuters first reported on the petition.
“We are all facing the impacts of what happens when worker protections, consent, and safety are ignored to scale at any cost,” protest organizers said in a statement shared with WIRED. “Together we can push back to ensure our rights are protected and our products are ethically built."
Notably, the tracking tool hasn’t been deployed outside the US because of what several workers described as stricter privacy rules and employee protections in other jurisdictions.
AI Panic
Meta is not alone in cutting jobs while spending lavishly on AI services and infrastructure. Block, Coinbase, and Cloudflare have all laid off thousands of employees, citing AI as a catalyst for the restructuring.
While Zuckerberg has publicly said he believes AI will amplify human abilities rather than replace them, he also acknowledged last month during an earnings call that AI is changing the pace of work. Projects that previously would have taken months and dozens of workers, he said, now require one or two people and could be finished in a week. Zuckerberg said that he’s building “the next evolution of our company around these people.”
Meta employees’ use of AI is tracked, and they receive data about their usage relative to groups of other colleagues. Vice presidents generally understand that their performance is judged partly based on driving automation in their units, according to someone who has spoken with one of the executives. Meta’s Clayton denies performance philosophies have changed, saying people are evaluated on their impact and have been asked “to embrace AI for themselves and their organizations to enhance productivity.” He says adopting AI is “a well-known” company priority and the company is “focused on using AI to help employees with their day-to-day work."
Some workers say they now feel pressure to automate sending emails or drafting reports, and worry that Meta is undermining one of its workplace slogans—not to mistake motion for progress. Executives have also urged some product teams to introduce generative AI features into Meta’s family of social media apps. “This is the part that’s triggered the real morale crusher,” a former employee says of the AI directives. “Zuckerberg was just obsessed that this is the transformation we have to go to.”
Some employees are on board. TBD Lab, which houses much of Meta’s top research talent and has been tasked with building frontier AI models, has been insulated from broader chaos and departures, according to an employee on the team. Both a longtime senior leader at Meta and an AI-pilled former engineer describe significant excitement across the company about generative AI systems as more and more employees are trained on them. “This is the opportunity of a lifetime,” the senior leader says of Meta’s workforce. “You have permission to go learn all of this and access frontier models, and you're surrounded by people who have this expertise.”
The person described Meta as making a big bet rather than playing it safe to “make $1 more” here or there. “There are going to be entire sections of the company where automation is better than the humans, and that's going to be really hard and sad, and those people are going to lose jobs, and [Zuckerberg] isn't sugarcoating that.”
Not everyone is worried about the impending job cuts, either. Some units such as policy have been told they won’t be affected this month, one employee says. Everyone else is in a “mad scramble” to finish projects and “to prove why they should be the ones spared,” they say. “It’s a lot to cram into a short period of time: Everyone ends up miserable.”
But the only official advice from HR to prepare for next week’s cuts was “make sure your personal email” is up to date internally—and wait.
Will Knight and Maxwell Zeff contributed to this report.