Ex-Employees Take Legal Action Against Meta for Claims of Discriminatory AI in Layoff Decisions
Twenty-six former employees of Meta have filed a lawsuit against the company, alleging that biased artificial intelligence tools were employed to disproportionately identify individuals on medical leave for layoffs. This claim follows Meta’s recent decision to reduce its workforce by approximately 8,000 employees in an effort to balance investments in AI and data centers.
The lawsuit contends that Meta utilized various AI systems, including an internal assistant named “Metamate,” staff-trained agents known as “second brain,” dashboards for tracking AI token usage, and data monitoring employee activities. These systems were allegedly intended to evaluate and rank employees based on performance, productivity, and engagement with AI tools. However, critics of the AI selection process assert that it failed to consider employees on protected family or medical leave and those whose disabilities could impact their productivity, particularly since such employees might be unable to utilize AI tokens or other digital resources.
In response to the lawsuit, Meta has expressed that the claims are unfounded, stating that human oversight was an essential component of their layoff decisions. “Workforce management and organizational choices were made by people, not AI,” the company asserted.
Central to the lawsuit’s argument is that all plaintiffs either took, requested, or were granted statutorily protected leave within two years prior to their layoff. Legislation like the Family and Medical Leave Act prohibits companies from factoring in protected leave when making employment decisions. Additionally, California’s Family Rights Act and the state’s Fair Employment and Housing Act restrict the use of automated systems that could lead to discrimination based on disability or gender, including pregnancy.
The plaintiffs are seeking to prevent Meta from proceeding with layoffs until an independent review of the algorithm-driven selection process is conducted. Due to the stipulations in their employment contracts, the plaintiffs also plan to advance their claims through arbitration.
Earlier reports indicated that Meta was documenting staff activities, including keystrokes and mouse movements, for AI training purposes. This practice raised concerns regarding privacy violations, particularly within the European Union. The company opted to suspend this AI training initiative when it became apparent that sensitive communications could be accessed by employees. Given the nature of these surveillance measures, the potential application of similar data in layoffs is not entirely surprising.
