Meta Employees Sue Over Alleged AI Bias in Layoff Selections

A group of 26 Meta employees has filed a federal lawsuit accusing the company of using AI assisted performance systems that allegedly placed workers with disabilities and those taking protected medical, parental, pregnancy, or family leave at a disadvantage during its latest layoffs. The anonymous plaintiffs filed the complaint in Oakland, California, and are seeking a preliminary court order blocking their scheduled terminations from beginning on July 22 while their claims proceed through private arbitration. All 26 plaintiffs remain employed by Meta at the time of filing.

According to the Reuters report, the complaint alleges that Meta used a collection of internal systems to score, rank, and select employees for termination. These reportedly included the Metamate large language model assistant, employee trained AI agents described as second brains, dashboards tracking AI token usage, and productivity measurements allegedly generated from communications and activity data. The plaintiffs claim these measurements failed to account for legally protected absences, meaning employees could accumulate lower productivity and AI usage scores simply because they were receiving medical treatment, managing a disability, caring for relatives, or taking parental leave.

The lawsuit argues that Meta’s selection process violated federal and state employment protections, including the Family and Medical Leave Act, Americans with Disabilities Act, Pregnancy Discrimination Act, and Pregnant Workers Fairness Act. The employees also claim that Meta failed to adequately test its automated systems for discriminatory bias. Their requested injunction is intended to preserve employment, health insurance, unvested equity, immigration status, and protected leave rights until their individual arbitration cases can be considered.

Meta strongly denies that artificial intelligence made the layoff decisions and says its workforce changes were controlled by human managers.

“These claims lack merit and are not based on facts. Workforce management and organizational decisions were and are made by people, not AI.
— Meta spokesperson”

The 26 employees were included in Meta’s May restructuring, which affected approximately 10% of its global workforce, or nearly 8,000 positions. CEO Mark Zuckerberg later told employees that he did not expect further company wide layoffs during 2026, although that statement did not reverse the terminations already announced as part of the May restructuring.

The dispute arrives as Meta dramatically increases its use of artificial intelligence across products, internal workflows, silicon development, and data center infrastructure. An internal memo reviewed by Reuters indicated that Meta plans to deploy 7 gigawatts of computing capacity during 2026 and reach 14 gigawatts in 2027. The company also expects to begin manufacturing its Iris AI accelerator in September with design support from Broadcom and production through TSMC. Meta could spend as much as $145 billion on AI infrastructure this year while continuing to deploy accelerators from NVIDIA and AMD.

The lawsuit does not yet prove that Meta allowed an algorithm to terminate employees autonomously. It does, however, raise a critical governance question about how AI generated productivity information influences human decisions.

A manager may technically approve the final termination while still relying heavily on rankings produced by automated tools. When those systems measure activity, output, communications, or AI usage without correctly excluding protected leave, they can transform legitimate absences into apparent underperformance.

Companies deploying AI for workforce management need transparent criteria, independent bias testing, human review, and clear safeguards for disability accommodations and protected leave. Without those controls, automation can reproduce discrimination at a scale and speed that traditional performance reviews could never achieve.

The case could become an important test of whether employers remain legally responsible when human decision makers rely on automated recommendations that produce discriminatory outcomes. Regardless of whether Meta ultimately prevails, the lawsuit shows that AI adoption inside corporate management is advancing faster than accountability frameworks designed to govern it.

Should companies be allowed to use AI generated productivity scores when selecting employees for layoffs, or should automated systems be prohibited from influencing termination decisions?