Meta has scaled back its controversial employee activity-tracking initiative after facing strong opposition from staff who raised concerns about privacy, surveillance and data collection practices. The company has now introduced new safeguards and greater employee control over the system, which was originally designed to collect workplace interaction data for training advanced artificial intelligence models.
The initiative, known internally as the Model Capability Initiative (MCI), was launched as part of Meta’s broader effort to develop AI agents capable of performing digital tasks autonomously. The software was designed to record mouse movements, clicks, keyboard activity and software interactions on employee work computers, providing real-world behavioral data to improve AI training.
However, the program quickly became a source of controversy within the company. Employees expressed concerns that the tracking system was too intrusive and collected excessive amounts of information from work devices. Critics argued that the initiative blurred the line between AI development and workplace surveillance, especially as Meta continues to invest heavily in automation technologies.
In response to employee feedback, Meta has introduced several changes to the program. Workers will now be allowed to temporarily pause data collection for up to 30 minutes at a time and can apply for exemptions under certain circumstances. The company has also optimized the software to reduce its impact on battery life and internet data consumption after employees reported unusually high resource usage.
According to internal communications, Meta maintains that the project underwent multiple privacy and risk assessments before deployment. Company officials insist that the collected information is intended solely for AI training purposes and not for employee performance evaluations. Nevertheless, staff concerns have remained significant enough to force a partial retreat from the original rollout plan.
The controversy comes at a crucial time for Meta as CEO Mark Zuckerberg accelerates the company’s artificial intelligence strategy. Meta has been aggressively investing in AI infrastructure, talent acquisition and next-generation models as it competes with rivals in the race to develop increasingly capable AI systems. The company believes that observing how humans interact with computers can help create AI agents that navigate software and complete tasks more effectively.
Privacy advocates have also raised concerns about the broader implications of the initiative. Some experts argue that collecting detailed behavioral data from employees could create regulatory challenges, particularly in regions with strict data protection laws. Earlier reports suggested that the system might inadvertently capture information involving non-U.S. individuals, potentially attracting scrutiny under European privacy regulations.
The employee backlash highlights a growing debate across the technology industry regarding the balance between AI innovation and workplace privacy. As companies seek larger datasets to train increasingly sophisticated AI models, questions surrounding consent, transparency and employee rights are becoming more prominent.
Industry analysts believe Meta’s decision to modify the program reflects a broader realization that successful AI development requires not only technical progress but also employee trust. While the company remains committed to using human-computer interaction data for AI training, the latest changes suggest that worker concerns can still influence the direction of major technology initiatives.
The episode serves as a reminder of the challenges facing technology companies as they push the boundaries of artificial intelligence. As AI becomes increasingly integrated into workplace systems, businesses will likely face greater scrutiny over how they collect, manage and use employee data in the pursuit of innovation.
