Meta Pauses AI Data Collection From Employees Amid Privacy Concerns

Meta has temporarily suspended a company-wide initiative to collect employee computer usage data for artificial intelligence (AI) training, following significant internal opposition and concerns over privacy.

The program, known internally as the Model Capability Initiative (MCI), began just two months ago. Its aim was to gather data on how employees used their computers, including mouse clicks and keystrokes, to refine AI models. However, it immediately faced strong criticism from employees who expressed discomfort with their every online action at work being tracked and questioned the security and destination of the collected data.

The pause was enacted on Monday after Meta realized that some of the gathered data had been potentially accessible to unauthorized personnel within the company. A Meta spokesperson confirmed to the BBC that the MCI program is "on pause for now" while the company conducts an investigation. The spokesperson added, "We have no indication at this time that any data was improperly accessed by Meta employees."

This decision comes after weeks of employee discontent, which included a petition signed by nearly 2,000 Meta workers calling for the cancellation of the MCI program. Initially, Meta attempted to address employee concerns by offering an opt-out period of up to 30 minutes at a time from the tracking, a measure described by one employee as a "damage control" attempt.

Another employee highlighted that while many technical staff are supportive of efforts to improve AI models and maintain competitiveness against rivals like Anthropic and OpenAI, the lack of consent and the mandatory nature of the tracking program fueled anger. "I've never seen morale here so bad," the employee stated.

Employee frustration at Meta has been exacerbated by recent extensive layoffs and a significant organizational shift towards AI initiatives, with the company reportedly investing up to $145 billion this year alone in AI. Reports from internal meetings have indicated open criticism of management regarding these AI-driven changes. A former employee described the company's direction as "exhausting and depressing."