Meta’s recent reduction in force allegedly targeted employees who had taken protected leave, according to a lawsuit filed by 26 current and former workers. The plaintiffs claim the company used artificial intelligence systems to score and rank staff for layoffs, and these tools disproportionately selected workers who had taken medical, pregnancy, or disability leave. A Meta spokesperson called the allegations “lack merit and are not based on facts,” stating that workforce decisions were made by people, not AI.
Company officials argue that the selection process was automated and did not account for the unique circumstances of staff members on medical or parental leave. Because the algorithms penalized employees for not producing data during their absence, those workers were selected for termination at higher rates than those who remained active. The legal team representing the workers cited a high-profile case involving Workday’s applicant-screening algorithms, which allegedly excluded workers over 40 and other protected groups.
Allegations of Systemic Bias
The lawsuit alleges that Meta’s internal AI systems rely on inputs that employees on protected leave cannot accumulate, such as performance ratings, calibration scores, and productivity metrics. The company reportedly did not adjust these inputs to account for time away from work. Because the algorithms penalized employees for not producing data during their leave, those workers were selected for termination at higher rates than those who remained active.
The workers describe a scenario where a scientist was chosen for a layoff while on pre-birth pregnancy leave. Another manager was demoted following a medical leave and then selected for termination weeks into a second medical leave. An engineer saw his rating lowered because of “broken time” caused by an injury. These examples highlight a pattern where the AI tools failed to recognize absences as temporary gaps rather than performance failures.
The legal challenge focuses on whether the automated selection process violated federal anti-discrimination laws, including the Americans with Disabilities Act, the Family and Medical Leave Act, the Pregnancy Discrimination Act, the Pregnant Workers Fairness Act, and Title VII of the 1964 Civil Rights Act. The plaintiffs are seeking a preliminary injunction to prevent the company from finalizing the separations.
It is difficult to compare this specific instance to past corporate layoffs, as the integration of AI into workforce management is a relatively new development. Historically, management had direct control over human resources decisions, allowing for individual discretion that is often absent in automated systems. When algorithms are used to filter candidates or rank employees, the resulting lists are often opaque, making it harder for individuals to understand why they were selected for termination. The plaintiffs argue that the lack of human oversight in this process allowed these systemic biases to go unchecked.
While Meta disputes the specific claims, the lawsuit aligns with a growing national conversation about algorithmic bias in employment decisions. Courts and regulators are increasingly scrutinizing how automated tools are used to screen applicants and manage workforces.
Experts suggest that companies must ensure their tools account for non-traditional work patterns. Moral arguments for fair chance hiring have gained traction, urging organizations to consider the full human context behind employment data. Without such considerations, automated systems risk perpetuating existing inequalities.
Leave a Reply