Key Takeaways
- Twenty‑six unnamed Meta employees allege the company used discriminatory AI‑driven tools to select workers for a May 2024 layoff that cut 8,000 jobs (10 % of the workforce).
- The lawsuit claims AI systems—including the internal large‑language model Metamate, productivity scores based on keystrokes, browser history, email data, and AI‑assisted performance reviews—penalized employees who took protected medical or family leave.
- Plaintiffs say the scoring “failed to account for their protected leaves, but in effect penalized the employees for exercising their legal rights to these leaves,” resulting in a disproportionate number of workers on maternity, paternity, or disability leave being chosen for termination.
- The employees seek a court order to halt the scheduled July 22 layoffs so they can pursue individual arbitration claims; Meta denies the allegations, stating decisions were made by people, not AI.
- The case follows a similar February 2025 lawsuit alleging age‑based discrimination in another round of layoffs, suggesting a pattern of contested workforce reductions at Meta.
Background of the May 2024 Layoffs
Meta announced a substantial workforce reduction in May that affected 8,000 employees, or roughly ten percent of its total staff. The company framed the cuts as a cost‑saving measure to fund its ambitious artificial‑intelligence initiatives, which it said would require “hundreds of billions of dollars” in investment. According to the complaint filed in the Northern District Court of California, the layoff list was not compiled through managerial judgment but rather through “a constellation of internal artificial‑intelligence systems” designed to “score, rank, and select employees for inclusion on the list.”
AI Systems Allegedly Used for Selection
The plaintiffs contend that Meta relied on several AI‑powered tools when deciding who would be let go. These included Metamate, an internal large‑language model assistant described as a “second brain” that had been trained on employee communications and documents. Additionally, the company allegedly used algorithmic productivity scores derived from keystroke monitoring, browser‑history tracking, and email‑volume metrics, alongside AI‑assisted performance‑review platforms. Records of internal AI token consumption were also said to factor into the scoring process. As the lawsuit puts it, “Meta did not assemble the termination list through the considered judgment of managers who knew the work.”
How the AI Metrics Disadvantaged Protected Leave Takers
Central to the claim is that the AI‑driven metrics penalized employees who had to miss work or reduce output because of a disability, medical condition, or protected family leave. Because the scoring models heavily weighted activity‑based data—such as keystrokes and email volume—workers on leave naturally received lower scores, even though their absence was legally protected. The lawsuit argues that this produced a discriminatory outcome: “The result was that employees who took protected leaves were disproportionately selected for layoff, based on scoring that not only failed to account for their protected leaves, but in effect penalized the employees for exercising their legal rights to these leaves.” Specific examples cited include a scientist who was two days away from giving birth and a manager on approved pregnancy‑related disability leave who was the sole member of her team flagged for termination.
Plaintiffs’ Legal Requests and Meta’s Response
The employees are asking the court to issue a preliminary injunction that would block Meta from completing the layoffs scheduled for July 22, thereby giving them time to pursue their claims through private arbitration, as required by their employment contracts. They argue that while Meta’s arbitration agreements cover individual workplace disputes, they do not extend to requests for temporary relief such as a halt to ongoing layoffs. Meta has denied the allegations outright. A spokesperson told Gizmodo, “These claims lack merit and are not based on facts,” and added, “Workforce management and organizational decisions were and are made by people, not AI.”
Connection to Prior Discrimination Allegations
This lawsuit arrives just months after another discrimination claim surfaced against Meta. In February 2025, a former employee alleged that older workers were disproportionately targeted in a round of layoffs that affected five percent of the workforce. At that time, Meta defended the cuts as targeting its “lowest performers.” Notably, one of the engineers involved in the current case said he was aware that employees who had taken paternity leave were also laid off during the February 2025 reductions, suggesting a possible pattern of bias across multiple workforce‑adjustment exercises.
Implications for AI‑Driven HR Practices
The case highlights growing scrutiny over how companies deploy algorithmic tools in human‑resources decisions. If the court finds that Meta’s AI systems inadvertently—or intentionally—penalized employees for exercising legally protected rights, it could set a precedent requiring greater transparency, auditing, and mitigation of bias in AI‑based performance and productivity metrics. Employers may need to reconsider reliance on raw activity data that fails to contextualize legitimate absences, and instead incorporate safeguards that account for leave, disability accommodations, and other protected circumstances.
Conclusion
The lawsuit against Meta underscores the tension between rapid AI adoption and the imperative to uphold fair labor practices. While Meta maintains that human managers—not algorithms—made the final layoff decisions, the plaintiffs’ evidence points to a system where AI‑generated scores heavily influenced those outcomes, disproportionately affecting workers on protected leave. As the case proceeds, its outcome could shape how tech giants and other industries balance efficiency gains from AI with compliance to anti‑discrimination laws and the ethical treatment of employees. The forthcoming court ruling on the requested injunction will be a critical first step in determining whether the alleged AI bias will be halted before the July 22 layoffs move forward.
https://gizmodo.com/meta-sued-for-allegedly-using-discriminatory-ai-in-layoff-decisions-2000785427