Dozens of employees at Meta have filed a lawsuit claiming that the company’s AI systems targeted them after they requested legally protected leave, maternity leave, or workplace accommodations due to disabilities, according to reports.
The lawsuit, initiated on Monday in a federal court in California, addresses layoffs that occurred earlier this year when Meta, the parent company of Facebook, Instagram, and WhatsApp, laid off approximately 8,000 workers.
The comprehensive 71-page complaint asserts that Meta employed a “constellation of internal artificial intelligence systems,” including automated assessments and tracking data, to determine which employees would be laid off.
According to the complaint, “Meta did not compile the layoff list based on a thoughtful decision from managers who understood the employees’ work.” Instead, the 26 employees named in the lawsuit allege that the company utilized AI systems “to evaluate, rank, and select employees for inclusion on the list.”
How Meta’s Algorithms Functioned
The legal action against Mark Zuckerberg’s company contends that the AI technology collected performance and productivity data that was unavailable during medical or family leave, and for individuals with disabilities, these metrics might be lower.
“The result was that employees who took protected leave were disproportionately selected for termination based on scores that not only ignored their protected absences but, in fact, penalized them for exercising their legal rights to take time off,” the complaint states.
Examples provided in the lawsuit include a researcher notified of her termination two days before giving birth, an engineer rated lower after being absent due to an injury, and a manager who was fired just 16 days after starting medical leave.
Zuckerberg’s Secret Monitoring Program
Earlier this year, Meta introduced a program designed to record employees’ keystrokes, mouse activity, browsing history, messages, and the location of company devices.
Zuckerberg internally outlined that the purpose was to train AI systems: “AI models learn by observing how truly intelligent people act.”
The complaint alleges that the program was launched secretly without employee consent, through an internal post with limited visibility. In June, Zuckerberg suspended the program after over 1,600 employees signed a petition against it.
Meta Denies the Allegations
As the use of AI in the workplace comes under increased scrutiny from regulators, states like California, Colorado, and Illinois have already passed laws to protect employees from biases inherent in automated decision-making systems.
A Meta spokesperson has denied the allegations, informing The Guardian that organizational decisions are made by people, not by AI.
The 26 plaintiffs, whose contracts expire on July 22, are seeking an independent audit of the AI tools and the maintenance of their employment status during the proceedings. Attorneys caution that if the layoffs are finalized, the damages could be irreversible, resulting in the loss of health insurance during pregnancy or medical treatment, as well as stock options held by the employees.
