Meta Listens to Staff, Scales Back AI Data Collection

When thousands of Meta employees spoke up about new workplace monitoring software, the company listened and made real changes.

Meta announced Tuesday it's scaling back its plan to collect employee mouse movements and keystrokes for AI training. The changes come after weeks of staff concerns about privacy, battery drain, and home internet spikes caused by the data collection software.

The company launched the monitoring program last month on U.S. employee computers as part of its push to build AI agents that can handle work tasks. The goal was ambitious: use real work patterns to train smarter artificial intelligence systems.

But employees pushed back hard. They raised concerns about personal data on work devices and complained that the software was draining laptop batteries and causing home internet bills to jump.

Meta's AI team, led by Vice President Stephane Kasriel, responded with concrete solutions. Workers can now pause data collection for up to 30 minutes at a time and request full exemptions from the program.

The team also optimized the software to reduce its drain on battery life and internet bandwidth. Kasriel acknowledged the feedback in an internal memo, saying the company heard concerns loud and clear despite being confident in its original privacy protections.

The Bright Side

This story shows workplace democracy in action at one of the world's largest tech companies. When employees united around shared concerns, management didn't dismiss them or push forward unchanged.

The outcome demonstrates that even in rapidly evolving fields like AI development, companies can balance innovation goals with worker comfort and privacy. Meta found a middle path that keeps its AI training program moving while giving employees meaningful control over their participation.

The changes prove that employee feedback matters, even when it slows down ambitious tech initiatives. Speaking up works, and companies benefit when they create space for honest dialogue about new workplace technologies.

This kind of responsive adjustment could set a helpful precedent as more companies explore AI training methods that involve employee data.