Twenty-six Meta employees selected for termination have sued the company, alleging that Meta used internal artificial intelligence systems and workplace surveillance data to decide who landed on an 8,000-person layoff list.
The complaint, filed Monday in the US District Court for the Northern District of California, says the process harmed employees with disabilities and workers who had taken protected medical or family leave. The plaintiffs filed as “Doe” employees, meaning their names are not public in the complaint.
The core allegation is blunt: Meta did not rely on managers making case-by-case judgments about employees’ work. The plaintiffs say the company used a set of internal AI and data systems to score, rank and select people for termination.
According to the complaint, those systems included an internal tool called “Metamate,” employee-trained “second-brain” agents, keystroke and activity monitoring data, dashboards tracking AI-token usage, and algorithmically assisted performance ranking and calibration.
That matters because the lawsuit says the systems did more than summarize performance records. The plaintiffs allege Meta’s internal tools helped generate the termination list itself, using measurements that could penalize workers for reasons unrelated to job quality.
AI usage allegedly became a workplace score
The complaint says Meta graded employees, in part, on how much they used the company’s AI tools. Internal dashboards allegedly sorted workers by their level of AI adoption, using labels including “AI Native,” “AI First” and “AI Enabled.”
The plaintiffs argue that tying layoff decisions to those metrics disadvantaged people whose disabilities, medical treatment, caregiving obligations or protected leave affected how they appeared in Meta’s systems. The filing alleges the result was a termination process that targeted workers in protected categories.
The complaint does not, by itself, prove that Meta broke the law. It is the plaintiffs’ account of how the company made layoff decisions, and Meta will have an opportunity to respond in court.
Still, the case lays out a specific version of a problem employers have preferred to discuss in mushier terms: what happens when companies use automated scoring systems inside personnel decisions, then treat the output as management judgment.
Here, the plaintiffs say the machinery included both productivity surveillance and adoption metrics for Meta’s own AI tools. In their telling, the layoff process rewarded workers who looked like heavier users of those tools and exposed others to termination through categories that managers did not meaningfully check.
The lawsuit asks a federal court to examine whether Meta’s use of those systems in the layoff process unlawfully harmed disabled employees and workers who exercised medical or family leave rights.
This story draws on original reporting from Ars Technica.
