A group of 26 Meta employees filed a federal lawsuit in Oakland, California, alleging the company used AI-powered tools and activity-monitoring data to select workers for mass layoffs earlier this year — a process they say disproportionately targeted employees who took medical, parental, or family leave or who have disabilities.
What the complaint alleges
The plaintiffs contend Meta relied on a suite of algorithmic inputs — including internal AI systems, keystroke and activity-monitoring metrics, token-usage dashboards and algorithmically assisted performance rankings — to identify employees for termination. According to the complaint, many of these metrics “by design, cannot be accumulated by an employee who is on protected medical or family leave, or whose output is reduced by a disability,” and the company did not account for protected leave when computing employee scores.
The workers, drawn from six states and Washington, D.C., say the company failed to pause automated scoring systems to perform individualized, leave- and accommodation-neutral reviews that federal and state law require. They seek a preliminary ruling to block Meta from completing the layoffs while their claims proceed in arbitration, arguing that termination before such relief would cause irreparable harms — such as loss of employer-sponsored health coverage during pregnancy or active medical treatment — that cannot be remedied by money alone.
Legal claims and statutes cited
The complaint asserts violations of several federal and state protections, including the Family and Medical Leave Act (FMLA), the Americans with Disabilities Act (ADA), the Pregnancy Discrimination Act, and the Pregnant Workers Fairness Act. The plaintiffs also allege Meta violated newly enacted local laws in places such as California and New York City that require bias testing or disclosures for automated decision systems.
Key federal references:
- Family and Medical Leave Act (FMLA)
- Americans with Disabilities Act (ADA)
- Pregnancy Discrimination Act
- Pregnant Workers Fairness Act
Meta’s response
Meta has publicly disputed the allegations, saying workforce decisions are made by people, not AI. In a statement included in news coverage of the filing, a Meta spokesperson said the claims “lack merit and are not based on facts,” and emphasized that workforce management and organizational decisions were made by humans. (Company statements and context about Meta’s automation policies are available through Meta’s public newsroom and corporate communications channels.)
For Meta corporate communications: Meta Newsroom.
Why this case matters
This lawsuit sits at the intersection of several fast-evolving areas: corporate use of automated monitoring and AI-assisted personnel decisions, employment law protections for employees on medical and family leave, and emerging regulation of automated decision systems. The case highlights at least three broader themes:
- Limits of activity metrics. Productivity and activity-monitoring tools (keystroke tracking, token-use dashboards, automated performance scores) can systematically undercount employees who are on leave or whose medical conditions reduce short-term output, creating a structural bias if those metrics feed automated selection processes.
- Regulatory and legal friction. New local and state rules increasingly require testing and disclosure for algorithmic systems, and existing anti-discrimination statutes may apply where automated processes have disparate impacts on protected classes or those using protected leave.
- Arbitration and interim relief. Many large employers use arbitration clauses requiring individual disputes to go to private arbitration. Plaintiffs here are seeking a court-ordered preliminary injunction to preserve employment status pending arbitration — a legal strategy designed to prevent irreversible harms (loss of health coverage, income, or parental leave benefits) before an arbitrator can hear the merits.
Potential implications for employers and lawmakers
If courts find that an employer’s use of algorithmic inputs in layoff decisions results in unlawful disparate impact or fails to provide legally required individualized review, companies may face injunctions, damages, and stricter oversight of their people-analytics tools. Employers using AI and automated monitoring may need to:
- Audit and validate models and metrics for disparate impacts on protected groups and on employees who take protected leave.
- Temporarily exclude leave periods from algorithmic scoring or institute manual review safeguards for employees affected by protected leave.
- Review arbitration agreements and consider carve-outs that allow immediate court review for requests for preliminary relief where irreparable harm is alleged.
Where this goes next
The plaintiffs are seeking a preliminary injunction to stop Meta from finalizing the layoffs while they pursue individual arbitration. The outcome will likely turn on whether the court finds plausible legal claims that the automated or algorithm-assisted selection process produced unlawful disparate impact or failed to provide legally required individualized review for employees on protected leave. A decision in favor of the plaintiffs could set precedent around interim relief in arbitration-covered disputes and increase scrutiny on employer use of monitoring technologies.
This lawsuit also adds to a broader public debate about the appropriate role of AI and automated monitoring in human-resources decisions and the safeguards necessary to protect workers who take protected leave or have disabilities.
Selected coverage and resources
- News coverage and reporting on the lawsuit: Fox Business (news reports quoted in the complaint reporting)
- Meta corporate statements: Meta Newsroom
- Federal statutes referenced: FMLA, ADA, Pregnancy Discrimination Act, Pregnant Workers Fairness Act
Note: This article summarizes allegations and positions as described in the plaintiffs’ complaint and reporting about it. The matter is pending in court; allegations in the complaint have not been proven. Meta has disputed the claims.
