Meta has implemented a tool called the Model Capability Initiative (MCI) to track employee activities on their computers, including mouse movements, keystrokes, and screenshots, to gather data for training its AI agents. While the company asserts that this data will not be used for performance assessments, it has faced internal backlash from employees concerned about privacy and the lack of an opt-out option.
Meta's "Model Capability Initiative" collects data from employee computer interactions, such as mouse movements and keystrokes, to train AI agents to better perform tasks as humans do. While this initiative aims to improve AI capabilities, it raises significant privacy concerns among employees, as participation is mandatory and lacks an opt-out option. For professionals interested in AI deployment and safety, this highlights the delicate balance between data-driven AI advancements and privacy considerations, suggesting that organizations need robust ethical frameworks and transparent communication to address such concerns.