Meta sued over AI-driven layoffs targeting disabled workers

On the heels of Meta's January 2024 layoff announcement affecting 8,000 employees, 26 workers have filed a federal lawsuit in the Northern District of California alleging that the company systematically used internal AI tools to select them for termination. The complaint, filed by plaintiffs listed as "Doe," claims Meta did not rely on the considered judgment of managers who directly knew the affected workers' contributions. Instead, it deployed what the lawsuit describes as "a constellation of internal artificial-intelligence systems," including a tool called Metamate, employee-trained "second-brain" agents, keystroke and activity-monitoring systems, AI-token-usage dashboards, and algorithmic performance ranking and calibration tools. These systems scored, ranked, and selected workers for layoff inclusion.

A central allegation is that Meta graded employees partly on their adoption of the company's AI tools. Internal dashboards classified workers into categories such as "AI Native," "AI First," and "AI Enabled" based on their stage of adoption. The lawsuit argues this framing is inherently problematic for certain protected groups. Employees on protected medical or family leave, and those whose output is reduced by disability, could not accumulate the productivity and AI-usage metrics that the algorithmic systems relied upon. The inputs the tools drew from—performance ratings, calibration scores, productivity metrics, "AI-native" classifications, and AI-token consumption—were "by design" unavailable to workers in these categories, the complaint states. This structural dependency on metrics that disabled workers and those on leave could not generate allegedly created a discriminatory effect, independent of any intent to discriminate.

Meta's response, provided to Ars Technica, states flatly that the allegations "lack merit and are not based on facts" and insists that "workforce management and organizational decisions were and are made by people, not AI." The company declined to comment further on the specific systems named in the lawsuit or the mechanics of how they were applied. According to Reuters, this is the first lawsuit against a major US company to directly challenge the use of AI in conducting layoffs, placing Meta at the center of an emerging legal question about employer accountability when algorithms play a role in employment termination.

Meta's use of algorithmic tools in its 2024 layoff of 8,000 employees has drawn legal scrutiny for the first time in a US lawsuit against a major technology company. The complaint names 26 plaintiffs and characterizes Meta's approach not as straightforward human decision-making informed by data, but as a reliance on a constellation of proprietary AI systems that scored and ranked workers based on performance metrics, productivity outputs, and adoption of Meta's own AI tools. The lawsuit's core claim is that this methodology created a structural barrier for employees in protected categories—those on medical or family leave, and those with disabilities—who could not generate the same quantitative signals that the AI systems prioritized. Meta's statement that 'people, not AI' made the decisions sidesteps the specific allegation: that the tools themselves were designed to produce a selection list that humans then approved, rather than humans independently evaluating each termination. The framing matters legally, as it touches on employer liability for algorithmic bias and the company's duty of care when deploying opaque systems in high-stakes employment decisions. The lawsuit's designation as the first of its kind suggests it may set a precedent for how courts evaluate the role of AI in large-scale workforce reductions.