Meta allegedly used employee activity to train AI, viral audio clip sparks concern

A viral audio clip allegedly from an internal Meta meeting has triggered controversy over claims that the company may be using employee activity data to train its artificial intelligence systems.

The recording, which has not been independently verified, appears to feature Meta CEO Mark Zuckerberg discussing how AI models can learn from observing skilled engineers, raising questions about the role of workplace monitoring in AI development.

According to claims circulating alongside the clip, Meta’s internal systems may track employee activity such as keystrokes, clicks, and on-screen behavior. However, there is no confirmed evidence that such data is being directly used for AI training.

The company has not issued any official response to the leaked audio, and its authenticity remains unverified.

The incident comes amid broader concerns in the tech industry over surveillance tools in workplaces and how companies use internal data to improve AI models. Critics argue that such practices blur the line between productivity tracking and the use of employee labour as training material for automation systems.

The controversy has also gained attention alongside ongoing restructuring and layoffs across the tech sector, where firms are increasingly shifting resources toward AI development.

While supporters of AI development argue that internal data helps improve model performance, the debate continues over consent, transparency, and the ethical limits of workplace monitoring in the AI era.