Meta sued by 26 employees who claim AI layoff tools penalised workers on medical or family leave

Twenty-six current and former Meta employees have filed a lawsuit accusing the company of using a suite of internal artificial intelligence tools to select workers for its May layoff round, allegedly discriminating against people who had taken protected medical, parental or family leave. The complaint, lodged in the US District Court for the Northern District of California in Oakland, targets the elimination of roughly 8,000 positions, about 10% of Meta's global workforce.

How the selection system allegedly worked

According to the lawsuit, Meta deployed what it calls a "constellation of internal artificial-intelligence systems" to score, rank and select employees for termination. The tools included an internal AI assistant known as "Metamate," employee-trained "second-brain" agents, keystroke- and activity-monitoring data, AI-token-usage dashboards, and algorithmically assisted performance rankings. Employees were graded on metrics such as productivity, output volume, performance ratings, AI-adoption scores and the number of AI tokens they consumed. Internal dashboards reportedly classified workers by their stage of AI adoption, using categories like "AI Native," "AI First" and "AI Enabled."

The result was that employees on protected leaves were disproportionately selected for termination based on a score that not only failed to account for their protected leaves, but in practice penalised employees for exercising their legal right to such leaves.

The leave penalty

Lawyers for the plaintiffs argue that the metrics used by Meta's systems could not be accumulated by anyone on medical or family leave, or whose output was reduced by a disability. The complaint states that Meta did not neutralise those inputs for protected leave, did not exclude leave-takers from the selection cohort, and did not pause the system for those employees. One employee said a manager warned them that taking leave would result in their selection for the layoffs. All 26 plaintiffs had requested, received or been approved for a statutorily protected leave, or had requested a reasonable accommodation for a disability, within the 24 months before the layoffs were announced.

Meta's response

Meta has rejected the allegations, calling them "without merit and not based on facts." In statements provided to several outlets, the company insisted that human judgment, not AI, drove the decisions.

Workforce management and organizational decisions were and are made by people, not AI.

The company did not offer further comment on the specific tools named in the lawsuit.

Legal basis and what the plaintiffs want

The lawsuit invokes the federal Family and Medical Leave Act, the California Family Rights Act, and the California Fair Employment and Housing Act, which forbids the use of automated-decision systems that produce disparate-impact discrimination on the basis of disability, sex or pregnancy. The plaintiffs are still employed by Meta, with employment terminations scheduled to begin on 22 July. They are asking the court to block the layoffs pending an independent audit of the algorithmically assisted selection process. Because of Meta's employment contract terms, the plaintiffs also intend to pursue their claims through individual arbitration.

Broader significance

Reuters has described the case as apparently the first against a major US company to challenge the alleged use of AI in conducting layoffs. The complaint puts a spotlight on the growing unease over how AI tools are deployed in personnel decisions, particularly when the data they rely on may systematically disadvantage workers exercising legally protected rights.