Meta sued by 26 workers alleging bias in AI-driven layoffs

Meta is being sued in federal court in Oakland by 26 current and former employees who say the company's AI-assisted performance scoring quietly pushed workers on medical, pregnancy, or parental leave onto its layoff list. CNBC has the filing, and it appears to be the first lawsuit against a major US technology company to directly challenge the use of AI in conducting layoffs.

The mechanics are what make this interesting. According to the complaint, Meta relied on factors such as productivity and AI token usage when it cut roughly 8,000 jobs, about 10% of its global workforce, beginning on May 20. The tools reportedly included Metamate, Meta's internal AI assistant, along with performance monitoring systems and token-usage dashboards that fed into scoring for layoff eligibility. If you were out on approved medical, pregnancy, or parental leave, you generated fewer tokens and less measured output, and the plaintiffs argue the system treated that absence as a negative factor rather than a protected one. Eight of the 26 are women who had taken maternity or pregnancy-related leave, four are men who had taken parental leave, and one is a woman who had taken leave to care for a family member and later bereavement leave.

The suit invokes the ADA, the FMLA, the Pregnant Workers Fairness Act, and the California Family Rights Act, and separately claims Meta failed to test its AI systems for bias in violation of recently adopted California and New York City laws. The immediate ask is narrow and sharp: a preliminary ruling blocking the plaintiffs' terminations, currently scheduled for July 22, while the underlying claims move into private arbitration. Meta's public line, as Fox Business reports, is that the suit lacks merit and that people, not AI, made the final workforce decisions.

The honest caveat is that these are allegations from an anonymous complaint at the earliest stage, and the reporting does not tell us how much weight Metamate's scores actually carried or whether human reviewers saw them at all. The forward-looking piece is that any company running an AI-assisted reduction in force now has a live legal case pointed at its playbook, and the eventual discovery here, if it survives the arbitration diversion, is going to be the thing every general counsel and head of HR reads next.