Twenty-six current and former Meta employees are suing the company over an AI-assisted layoff process that allegedly punished workers for being absent while on approved medical or family leave.
The irony is hard to miss: tools meant to measure productivity may have treated legally protected leave as poor performance.
TL;DR
- Twenty-six workers allege Meta used AI-assisted systems and productivity metrics to select employees for layoffs.
- The plaintiffs say approved medical, family and pregnancy-related absences were not properly considered.
- They want a federal court to pause layoffs scheduled to begin July 22.
- Meta says people, not AI, made the workforce decisions.
Meta Employees Say AI Layoff Rankings Penalized Medical And Family Leave
The lawsuit, filed in federal court in Oakland, California, claims Meta relied on a “constellation of internal artificial-intelligence systems” when determining which employees to cut.
According to the complaint, the systems drew on performance ratings, calibration scores, productivity data, “AI-native” ratings and AI-token consumption.
The workers argue that these metrics inherently disadvantaged employees who could not accumulate the same level of output because they were on protected medical or family leave, managing a disability or dealing with pregnancy-related absences.
The plaintiffs allege that the process violated federal and state protections against disability discrimination, leave-related retaliation and pregnancy discrimination.
The complaint also says Meta failed to test its AI systems for bias under recently adopted laws in California and New York City. The anonymous plaintiffs are based across six states, including California and New York, as well as the District of Columbia.
Meta’s Internal AI Tools Take Center Stage In Layoff Dispute
The lawsuit identifies several tools that allegedly contributed to employee scoring and rankings.
These include “Metamate,” Meta’s large language model assistant, an employee-trained “second brain” that tracked communications and documents, and a productivity score allegedly generated from keystrokes, screen content, emails and browser history.
AI-token consumption, which can act as a proxy for how frequently employees use AI tools, is another disputed input. The workers claim that employees on approved leave had fewer opportunities to generate the activity these systems rewarded, creating a structural disadvantage in the scoring process.
Lawsuit Seeks To Block Meta Layoffs Beginning July 22
The 26 plaintiffs were notified in May that their roles would be eliminated starting July 22. The notices were part of a layoff round affecting about 10% of Meta’s global workforce, or nearly 8,000 people.
The workers are asking the court for a preliminary injunction that would preserve their employment while they pursue individual claims in private arbitration. Their request also seeks an independent audit of Meta’s algorithmically assisted selection process.
The case appears to be the first major U.S. lawsuit directly challenging the alleged use of AI in conducting layoffs. That could make it an early test of how existing employment protections apply when automated systems help shape workforce decisions.
Topics For More Insights
- All You Need To Know About Meta AI's New Search And Editing Tools
- The Tech Hunger Games: Meta Improves Snack Perks Amid Low Morale; Elon Musk’s X Does Better
- Meta Pauses AI Training Program Tracking Employee Keystrokes After Internal Data Leak
- Meta’s Legal Showdown In New Mexico Could See The Social Media Giant Exit The State
Meta Denies AI Made Workforce Decisions As Legal Scrutiny Grows
Meta has rejected the allegations. A spokesperson said the “claims lack merit and are not based on facts.”
“Workforce management and organizational decisions were and are made by people, not AI,” the spokesperson said.
The lawsuit follows separate legal scrutiny involving Workday, where a federal judge recently allowed discrimination claims related to AI-powered hiring tools to proceed. Together, the cases raise a growing legal question for employers: even when people make the final decision, companies may still face challenges if the systems informing those decisions produce discriminatory outcomes.
