A group of Meta employees is suing the social-media company over how it carried out layoffs this spring, alleging that it used AI that disproportionately cut workers with disabilities or on protected leave.
The federal suit, filed this week in the Northern District of California, alleges that when Meta began laying off 8,000 employees—or roughly 10% of its workforce in May—it relied on a “constellation of internal artificial-intelligence systems” to create a termination list.
The suit claims the AI used metrics such as productivity and AI-token use to determine layoffs, and so targeted employees who had missed work because of medical conditions or other leave—violating discrimination laws in the process.
The AI tools “draw on inputs—performance ratings, calibration scores, productivity and output metrics, ‘AI-native’ ratings, and AI-token consumption—that, by design, cannot be accumulated by an employee who is on protected medical or family leave, or whose output is reduced by a disability,” the suit states.
A Meta spokesman rejected the allegations and said that workforce-management decisions were made by people, not AI. “These claims lack merit and are not based on facts,” he said.
The suit is one of the first in the U.S. to scrutinize the potential role of AI in determining layoffs. It also raises questions about the extent to which workers are shielded from potential cuts while on legally protected leave.
The plaintiffs in the suit are 26 current and former Meta employees, all of whom were notified they would be laid off. They had all either taken or requested leave or asked for disability-related accommodation in the months before.
One was on preapproved pregnancy leave, and was told she was being laid off two days before giving birth. Another was on medical leave for a “serious health condition,” the suit claims.
The plaintiffs are seeking a preliminary injunction to maintain the status quo of their employment pending arbitration, and to prevent Meta from completing their separations.
A leave of absence doesn’t automatically safeguard someone’s job, employment lawyers say, though companies also can’t single out those on a protected leave.
“It’s not a magic bullet,” said Jeffrey M. Hirsch, a professor of law at the University of North Carolina, who studies employment issues. “On the other hand, it does mean if your employer is picking and choosing who gets laid off, they can’t say that I’m annoyed so-and-so keeps taking FMLA leave,” he added, referring to the Family and Medical Leave Act.
Hirsch said the suit showed how the alleged use of new AI tools could be exposing companies to new legal risks.
The lawsuit also touches on a bigger question that has consumed white-collar employees for years: How exactly do companies choose who to include in a layoff? Inside many companies, divisional leaders and department heads often decide who should be eliminated, human-resources executives say, but creating a termination list is often messy and chaotic.
Meta announced the layoffs as part of a broad effort to reimagine its workforce and become more nimble to compete with AI-native startups. In May, CEO Mark Zuckerberg thanked laid off employees for their contributions and called AI “the most consequential technology of our lifetimes.”
The suit alleges that Meta used a range of AI-assisted systems to score, rank and select employees to include on its layoff list. Those included AI agents that “ingest each employee’s communications and documents to replicate the employee’s output,” the suit claims, along with productivity-monitoring tools and factors such as each employee’s rolling 12-month performance rating.
Write to Chip Cutter at chip.cutter@wsj.com and Meghan Bobrowsky at meghan.bobrowsky@wsj.com