Meta laid off thousands of employees to become an "AI-first company" in May, and some of those former employees are now alleging that Meta used AI to illegally fire them. Twenty-six former Meta employees have filed a lawsuit alleging that Meta's termination process was heavily aided by AI and violated worker protection laws.
Meta laid off 10% of its workforce, around 8,000 workers, two months ago. The lawsuit, filed on Monday in a US District Court in Oakland, California, alleges that Meta used algorithmic and AI tools to monitor specific employee "productivity metrics," and low performers were flagged as likely targets for layoffs.
However, many of those flagged employees had either approved accommodations or had taken legally protected leave in the past two years for pregnancy, medical reasons or bereavement, according to the lawsuit.
"Meta did not assemble the termination list through the considered judgment of managers who knew the work," the complaint alleges. "Instead, Meta used a constellation of internal artificial-intelligence systems … to score, rank, and select employees for inclusion on the list."
The result was that "employees who took protected leaves were disproportionately selected for layoff," the lawsuit alleges.
Under California state and federal employment protection laws, including the Americans with Disabilities Act and the Family and Medical Leave Act, employers are barred from retaliating against employees who need disability accommodations or use protected leave.
The complaint alleges Meta used tech and AI systems to monitor and select which employees to fire. These tools included "Metamate," an internal AI assistant; employee-trained agents built using an employee's digital communications; and algorithmic productivity scores that account for employees' keystroke logs, browser history and screen content. The plaintiffs claim that AI token use, monitored in an internal leaderboard, was also taken into account.
The former employees allege that because they weren't generating enough data for Meta's internal monitoring systems, their scores were dinged. They say this made it more likely for the AI to flag them to management when it came time for layoffs, despite having average and above-average manager reviews.
"Numerous other employees noticed that their own scores had dropped dramatically upon returning from vacation and protected leaves," the complaint alleges, "and that there was no effort made by Meta to '' a person's scoring when they were away from work based on excused, and even legally protected, absences."
A Meta spokesperson told CNET that these allegations "are not based on facts" and that "people, not AI," are responsible for workforce and organizational decisions.
It was reported in April that Meta was using AI to monitor its employees, including their AI token usage and keystroke logs. Employees reportedly felt spied on, and Meta scaled back but didn't reverse the policy in June.
This case highlights the growing role of AI in managerial decisions, including hiring, performance reviews and firing. But AI and algorithmic-based systems have a history of replicating institutional biases against workers with disabilities, along with minorities. A recent Harvard University study found significant racial disparities in AI job applicant screening software, leading to fewer Black and Asian candidates being recommended for interviews, for example.