Meta Faces Lawsuit Alleging AI-Based Layoff Process Discriminated Against Employees With Medical Conditions

Twenty-six former employees have filed a lawsuit against Meta Platforms in a federal court in Oakland, California, claiming the company used AI-supported systems during its latest round of layoffs in a way that unfairly affected workers with disabilities, employees who took medical leave, and those caring for family members.

According to the complaint, Meta relied on measures such as productivity data and AI token usage when deciding which employees would be laid off. The plaintiffs argue that these criteria placed workers who were absent because of medical or family-related reasons at a disadvantage.

The employees, who were informed in May that their jobs would end beginning July 22, are asking the court to temporarily stop the layoffs while they pursue their claims through private arbitration. They state that although workplace disputes are generally subject to arbitration under their employment agreements, requests for temporary court relief are not.

Meta rejected the allegations. A company spokesperson said the claims have no merit and stated that workforce and organizational decisions were made by people rather than artificial intelligence.

The lawsuit is believed to be the first legal challenge against a major U.S. company over the alleged use of AI tools in making layoff decisions.

Earlier this year, Meta announced job cuts affecting around 10% of its global workforce, or nearly 8,000 employees. The company has been restructuring while increasing its investment in artificial intelligence. CEO Mark Zuckerberg has since said he does not expect any additional company-wide layoffs this year.

The lawsuit also alleges that Meta violated federal and state laws protecting employees with disabilities, workers on medical leave, and pregnant employees. It further claims the company did not evaluate its AI systems for potential bias as required under recently adopted laws in California and New York City.

The plaintiffs, who remain anonymous, are from California, New York, four other states, and the District of Columbia. The complaint says Meta used several internal AI-assisted tools to evaluate employees, including "Metamate," an internal large language model assistant, an employee-developed knowledge system, and a productivity score generated from workplace activity such as keystrokes, screen content, emails, and browser history.