Meta Employees Sue, Claiming AI Flagged Workers on Leave for Layoffs
A federal lawsuit alleges Meta used automated performance tools and activity monitoring to target staff for redundancy after they took maternity or disability leave.
Key points
- Dozens of Meta employees filed a federal lawsuit in California on Monday alleging the company used artificial intelligence tools to select them for layoffs.
- All plaintiffs say they were targeted after requesting maternity leave, disability leave, or other legally protected time away from work.
- The lawsuit connects the firings to Meta's 2025 reduction of roughly 8,000 jobs.
- The suit names a "constellation of internal artificial intelligence systems," including AI-generated performance scores and software that logged keystrokes and computer activity.
Meta, the company that owns Facebook, Instagram, and WhatsApp, is facing a significant legal challenge from its own former workers.
Decades of employees filed suit in federal court in northern California on Monday, alleging that Meta used artificial intelligence tools, meaning automated software that scores and sorts employees without a human reviewing each decision, to decide who to cut during a major round of job losses earlier this year. Meta shed roughly 8,000 workers in that reduction.
The core claim is straightforward: the workers say they were flagged for dismissal because they had taken leave protected by law. That includes maternity leave, taken by new parents, and disability accommodation, meaning time or adjustments granted to employees with a medical condition.
According to the suit, first reported by The Guardian, Meta relied on what the plaintiffs call a "constellation of internal artificial intelligence systems." That phrase covers two distinct tools. The first is an AI-generated performance rating, a score assigned to each worker by software rather than by a manager sitting down and evaluating them. The second is activity-monitoring software that tracked keystrokes and general computer use.
When employees take protected leave, their activity data drops. Fewer keystrokes, less logged screen time. The lawsuit argues the AI systems read that drop as poor performance, then marked those workers for layoffs, without any human stepping in to question the result.
Meta has not yet responded publicly to the lawsuit in detail.
What does this mean for ordinary workers?
This case matters beyond Silicon Valley. If the allegations hold up in court, it would mean that automated workplace tools, now used at many large companies, can quietly penalise workers for exercising legal rights, without any manager ever consciously deciding to do so.
For anyone in a job that uses performance-monitoring software, the practical takeaway is to keep a record. Document when you requested leave and what your employer said. If your performance rating changes during or after protected leave, ask for an explanation in writing.
The lawsuit also raises a broader point. Employers in many countries are legally required to ensure that automated decision-making does not discriminate against protected groups. A system that scores a parent on maternity leave the same way it scores someone who never took a day off is not neutral, whatever its designers intended.
The case will now move through the federal court system. A ruling could take years, but it arrives at a moment when AI-driven HR tools, software that helps companies hire, rate, and fire staff, are spreading rapidly through workplaces of every size.
