Meta Employees Sue Company Over AI-Selected Layoffs

Twenty-six employees of Meta Platforms filed a class-action lawsuit against their employer on Monday, accusing the company of systematically eliminating workers on medical, parental, and caregiving leave during the AI-assisted layoff wave in May. The case was filed in federal court in Oakland, California.

According to court documents, Meta relied on internal AI systems, keystroke and activity monitoring data, dashboards tracking token usage in AI tools, and algorithmically assisted performance rankings to select employees for layoffs. The plaintiffs argue that some of these metrics cannot, by design, rise for an employee on protected medical or family leave, or for someone whose productivity drops because of a disability.

How the algorithm worked

The systems analyzed, among other things, annual performance reviews, so-called AI-native ratings, the number of tokens used in AI tools, and the content of communications and documents produced by employees. Lawyers representing the plaintiffs argue that this method of measurement inherently penalizes people who are absent from the office, regardless of the reason for their absence.

The lawsuit includes specific examples. One woman was reportedly laid off two days before giving birth, despite having previously approved maternity leave. Another employee was undergoing treatment for a serious medical condition at the time of the layoff. The plaintiffs' lawyers stress that the consequences of such decisions are irreversible.

once these separations are final, the harms are irreversible: employer-subsidized health coverage lost during pregnancy, postpartum recovery, and active medical treatment; time-bound leave rights extinguished; unvested equity forfeited; and immigration consequences triggered - from the lawsuit against Meta Platforms

Meta's response

A Meta spokesperson rejected the allegations, saying that personnel decisions were made by people, not AI systems.

the claims lack merit and are not based on facts. Workforce management and organizational decisions were and are made by people, not AI - Meta Platforms spokesperson

Gender and disability concerns

The plaintiffs point out that the algorithmic practice of recording absence as reduced productivity hits women harder than men, since women more often take longer leaves related to pregnancy and childbirth. A lawyer cited by US media notes that such a mechanism can violate anti-discrimination law even when the algorithm itself has no direct knowledge of an employee's gender or health status, since it is enough for it to correlate with those traits through absence patterns.

What it means for companies

Meta's case illustrates a practical problem facing companies rolling out AI for workforce management at scale: an algorithm optimized for productivity and presence metrics can, in practice, reproduce discrimination even if no one programmed it to do so deliberately. US labor law protects employees on medical and parental leave regardless of whether the layoff decision was made by a human or an automated system, and using an AI tool does not relieve an employer of liability.

For Polish companies considering similar tools for selecting layoff candidates, the case is a warning: the EU's AI Act (the bloc's artificial intelligence regulation) classifies systems used to evaluate and dismiss employees as high-risk, which comes with additional documentation and audit obligations. The Oakland court's ruling could become a reference point for future disputes over algorithmic layoffs, including outside the United States.

The case is at an early stage, and Meta will have to formally respond to the allegations in federal court. The company has not disclosed whether it will delay or modify the layoff wave planned for July 22 in light of the lawsuit.