Meta is making small adjustments to its controversial employee tracking program after facing internal pushback from workers. The social media company reportedly told employees they can now pause the monitoring system for up to 30 minutes when they need to “check something personal.”
The changes to the Model Capability Initiative (MCI) represent a minor retreat from Meta’s original plan to track nearly every keystroke and mouse click from its workforce. However, the vast majority of employees will still be required to participate in the AI training program, which records their computer activities to help improve Meta’s artificial intelligence models.
The tracking program highlights the growing tension between Big Tech companies’ AI ambitions and employee privacy concerns. As tech giants race to develop more sophisticated AI systems, they’re increasingly turning to their own workforces as data sources. Meta’s approach is particularly aggressive, essentially treating employees as unwitting trainers for AI models that could eventually automate parts of their jobs.
Beyond the 30-minute pause feature, Meta will allow a limited subset of workers to opt out entirely. This exemption applies only to:
- Remote workers with bandwidth limitations
- Employees handling sensitive materials
- Workers who frequently operate in spaces without reliable laptop charging access
The company also improved the software’s battery efficiency after employees complained about performance issues. These tweaks come after Meta faced internal protests when it announced MCI last month, coinciding with the layoff of 8,000 workers and a major reorganization that shifted thousands of employees into AI-focused roles.
CEO Mark Zuckerberg defended the program during a recent company-wide meeting, arguing that “watching really smart people do things” offers the best path for rapid AI improvement. He emphasized that Meta employees possess “significantly higher” intelligence than the average population, making their work patterns valuable training data.
Zuckerberg insisted the program isn’t designed for surveillance or performance monitoring. “None of the data is being used for, like, looking at what people are doing, or surveillance, or performance track[ing], or anything like that,” he said in leaked audio. Instead, he positioned it as feeding “a very large amount of content into the AI model, so that way it can learn how smart people use computers to accomplish tasks.”
The Meta CEO also hinted at expansion plans, suggesting the company would “probably do more things like it” if the current program proves successful. This raises questions about how far tech companies will go in harvesting employee data for AI development, and whether other major firms will adopt similar tracking measures.
The controversy reflects broader concerns about workplace privacy in an era where companies view employee behavior as a potential goldmine for AI training. While Meta frames MCI as a technical necessity for AI advancement, critics see it as a concerning erosion of worker autonomy and privacy rights.