Meta Platforms faces mounting European privacy concerns over a new internal tool designed to train AI models. The Model Capability Initiative (MCI) collects employee computer usage data to build AI agents for workplace automation.
The program tracks U.S. employees' mouse movements and clicks across more than 200 applications. Internal documents reveal the tool incidentally captures communications with staff located in the European Union.
Regulators are scrutinizing the program for potential violations of the EU’s General Data Protection Regulation (GDPR). This regulatory clash could result in significant financial penalties and jeopardize a core component of Meta’s AI strategy.