Meta Sued by 26 Former Employees, AI Allegedly Involved in Determining the List of Layoffs

JAKARTA - A total of 26 former Meta employees sued the company. They accused the artificial intelligence system of determining which workers were on the list of layoffs or layoffs.

CNET quoted Tuesday, July 21, saying the lawsuit was filed on Monday in the United States District Court in Oakland, California.

The plaintiffs assessed that the process was detrimental to workers who took medical leave, maternity leave, bereavement leave, or needed work adjustments due to disability.

Meta has previously cut around 10 percent of its workforce or around 8,000 people. The move comes as the company pushes to transform into a company that focuses more on AI.

According to the lawsuit, Meta uses algorithmic and AI tools to monitor employee productivity. Workers with low scores are then more easily included in the layoff list.

However, many workers who are marked as system employees have taken legally protected leave in the past two years.

"Meta does not draw up a list of dismissals through the careful consideration of managers who understand the job," the lawsuit reads.

"Instead, Meta uses a series of internal artificial intelligence systems to score, rank, and select employees for inclusion on the list."

The lawsuit alleges that workers who took protected leave were disproportionately selected for termination.

The former employees also claimed that their scores dropped because during the leave they did not produce enough data for Meta's monitoring system. In fact, some claimed to still receive an average or above average rating from managers.

According to the lawsuit, Meta did not stop the assessment during workers' leave or absence for legitimate and legally protected reasons.

In California and at the federal level, companies are prohibited from punishing workers who use protected leave rights or need work adjustments due to disabilities.

The protection is among others regulated in the Americans with Disabilities Act and the Family and Medical Leave Act.

The system in question includes Metamate, Meta's internal AI assistant, an AI agent trained using each employee's digital communications, as well as productivity scores based on keystroke records, internet browsing history, and screen content.

The use of AI tokens monitored through an internal ranking board is also said to be taken into account. AI tokens are small units in the form of words or word fragments that are processed when someone uses an artificial intelligence system.

Meta denied the allegations.

A company spokesperson told CNET that the allegations were "not based on facts". Meta also stressed that decisions on staffing and organization are made by humans, not AI.

In April, Meta reportedly used AI to monitor employees, including the use of AI tokens and keystroke records. A number of workers claimed to feel watched.

Meta later reduced the enforcement of the policy in June, but did not completely stop it.

CNET also quoted a recent study from Harvard University that found racial disparities in AI-based applicant screening software. The system recommended fewer black and Asian candidates for interviews.