Meta Moves Forward With Its Controversial Plan to Track Employees — But There’s a Way Out

Meta told workers a few months ago that it would track their clicks and keystrokes to train AI. After weeks of internal backlash, Meta revealed in an internal memo obtained by Reuters earlier this week that it decided to offer partial pauses and exemptions for those who object to the surveillance system.

“While we remain ​confident in ​the ​privacy protections we ‌put in place at launch, which went through several layers of risk review, we have heard your ‌concerns about ​personal data on ​work ​devices, battery life ‌and wanting more control ​over ​when capturing happens,” the company said in the ​memo.

According to Reuters, new controls will permit employees to pause the data collection for “up to 30 minutes at a time.” Workers can also request to be removed from the initiative entirely. For employees who do not want their activity tracked, these pauses and possible opt-outs now represent a way out. 

Meta introduced the Model Capability Initiative in April

In April, Meta briefed staff on a new internal tool that would monitor how employees use workplace software on company devices. The system, called the Model Capability Initiative (MCI), logs keystrokes, mouse movements, clicks and navigation through menus to generate real-world training data for Meta’s AI agents. 

Meta told the BBC at the time that the goal was to create computer agents to help people complete routine tasks. That meant their models needed “real examples” of how people actually use apps. The company stressed that the captured data would “not be used for any other purpose” and said the system was equipped with safeguards to protect sensitive data on work devices. 

Meta framed MCI as an extension of existing monitoring on corporate devices rather than a radical break, noting that activity on company hardware has long been accessible under standard policies. The distinction, however, was that Meta would dedicate this new layer of tracking to training and refining AI systems, instead of using it for routine security or compliance purposes. 

Immediate employee pushback

The internal reaction was swift and negative. Staff expressed alarm that Meta would use their everyday computer behavior to train AI models, especially after recent Meta layoffs linked to automation and efficiency. One employee told the BBC the plan felt “very dystopian,” adding to fears that tools trained on their work might one day help eliminate their jobs.

Another person who recently left Meta told the BBC that the tracking system was “just the latest way they’re shoving AI down everyone’s throat.” 

Within weeks, more than 1,500 Meta employees signed a petition opposing the initiative and urging Meta to rethink the program. They cited concerns such as privacy and data security, as well as battery drain on laptops and a lack of control over exactly when the software was watching. Several employees found that the tracking system took up so much data that it was causing their Internet use to surge when they worked from home. 

Meta responded to these pressures by offering the surveillance pauses and opt-outs. Meanwhile, in May, Meta also reduced its headcount by 10%, letting go of 8,000 employees.