Meta accused of using AI to target employees on medical leave

Meta Faces Allegations of Misuse of AI

Meta, the social media giant, is at the center of a legal controversy involving 26 of its former employees. The plaintiffs have filed a lawsuit, alleging that the company used artificial intelligence systems to determine which employees should be laid off. According to the complaint, this method particularly disadvantaged those who had taken medical, parental, or disability leave.

The plaintiffs argue that Meta utilized AI tools to assess employee performance based on surveillance data and productivity metrics. This approach allegedly led to the elimination of around 8,000 positions in May, according to documents filed in a federal court in Northern California. Employees on leave reportedly could not accumulate the necessary points according to these criteria, placing them in a position of underperformance in the eyes of the company.

Controversial Selection Methods

The complaint details that Meta allegedly used various technological tools to conduct these evaluations, including an internal chatbot named Metamate, AI usage dashboards, and virtual agents referred to as "secondary brains." These tools, powered by surveillance data of employee activity, purportedly allowed for ranking workers based on their adoption of AI, with categories such as "AI Native," "AI First," and "AI Enabled." Employees absent due to leave would have seen their scores decrease, placing them at the bottom of the performance scale.

The allegations have not yet been examined by a court, but Meta has already responded through a spokesperson. The spokesperson stated that the accusations were unfounded and not based on concrete facts. They also claimed that decisions regarding workforce management were made by people, not automated systems.

AI at the Heart of Meta's Strategy

This case comes at a time when Meta is actively encouraging its employees to integrate AI more into their daily tasks. The company has implemented internal dashboards to track the adoption of these technologies and has established rankings to evaluate this integration. Other major companies, such as Disney, JPMorgan, and Visa, have also begun measuring their employees' use of AI.

Among Meta's internal initiatives, an AI program was introduced this year, involving the installation of software on employees' computers to capture keystrokes and mouse movements. This data was intended to train the company's AI models but raised concerns among employees regarding privacy and the mandatory nature of the program. The complaint indicates that this data was also used to feed the layoff selection systems.

Implications and Plaintiffs' Demands

Following an internal leak in June, Meta suspended this surveillance program. Additionally, the company introduced a new performance system called Checkpoint, which emphasizes results and offers larger bonuses to the highest-rated employees. The complaint highlights that an AI-assisted element of Checkpoint made AI adoption a central evaluation metric.

The plaintiffs assert that several employees were targeted for layoffs while they were on approved leave or shortly after their return. One engineer stated that their manager had lowered their rating due to an injury that prevented them from working. Another employee reported that a manager warned them that taking medically approved leave could lead to a "permanent" inclusion on the layoff list.

The employees are asking the court to suspend their layoffs while their claims are reviewed in arbitration. They also seek an independent audit of Meta's layoff selection process and for the selections to be recalculated without considering protected leave or accommodations for disabilities.

The plaintiffs, who have chosen to remain anonymous, worked for Meta in several U.S. states, including California, Washington, New York, Illinois, Pennsylvania, and Florida.

The Reuters news agency was the first to report on this matter.