AINEWS
Twenty six Meta workers allege AI performance scores and keystroke monitoring turned protected leave into a layoff signal, and their federal complaint asks a judge to halt the 8,000 person cuts.
A 26-worker federal complaint against Meta alleges the company's internal AI systems did more than rank workers for an 8,000-person layoff list. The 71-page filing, lodged Monday in the Northern District of California, claims the systems converted the act of taking protected leave into a negative performance signal: the legal right not to be at work became, in the algorithm's output, a reason to be flagged for termination.
The named plaintiffs are evidence of a structural problem, the complaint argues, not isolated cases. A scientist was notified of her layoff two days before giving birth. An engineer saw his performance rating drop after injury time off. A manager was let go 16 days into medical leave. The plaintiffs describe a "constellation of internal artificial intelligence systems," including AI performance ratings, keystroke and activity monitoring, AI token-usage dashboards, and algorithmically assisted performance rankings, working in concert to score employees for the termination list during this year's approximately 8,000-person layoff.
Meta pushed back on the allegations. A company spokesperson said the claims "lack merit and are not based on facts" and insisted that "workforce management and organizational decisions were and are made by people, not AI." That denial sits against a complaint that describes a much more automated process. The keystroke, mouse, browser-history, message, email, and location monitoring program was rolled out earlier this year via what the filing calls a "low-visibility internal post" by an engineer rather than a senior leader, with some teams receiving no consent prompt at all, per the complaint.
The lawsuit is also a test of a broader legal question: do existing federal leave protections and the newer state "automated decision system" rules reach a system that converts protected leave into a performance penalty? California, Colorado, and Illinois have all enacted laws or regulations targeting AI-related employment bias in recent years, and the plaintiffs' disparate-impact allegations map onto the gap those statutes are designed to close. At the federal level, the Family and Medical Leave Act and the Americans with Disabilities Act supply the baseline anti-discrimination framework, even though neither law was drafted with algorithmic scoring in mind.
The complaint lands against internal messaging in which Meta leadership has defended the company's AI workforce tools in blunt terms. In an internal meeting reported by The Information and cited in the Guardian's coverage, Mark Zuckerberg framed the AI monitoring program this way: "The AI models learn from watching really smart people do things. The average intelligence of the people who are at this company is significantly higher than the average set of people that you can get to do tasks." The plaintiffs' theory is that the same systems Zuckerberg describes as identifying high performers also produced a list in which pregnancy, injury, and medical leave counted against workers.
The 26 named plaintiffs are a small subset of the roughly 8,000 workers Meta cut this year, and the legal burden on them to show disparate impact across that larger pool is steep. The plaintiffs are asking the court for a preliminary injunction that would freeze the layoffs before they finalize, alongside relief including reinstatement, back pay, lost equity, benefits, and damages. A first hearing in the Northern District of California will set the schedule for whether the algorithm-built layoff list survives judicial review.