Meta will capture employee keystrokes, mouse movements, and clicks to train artificial intelligence systems, according to a Reuters report confirmed by the company.
Meta's communications department confirmed the practice to Engadget, stating the data collection will support AI development efforts. The initiative marks an expansion of data sourcing for machine learning, following Meta's use of publicly available internet content and published books for AI training.
The scope of data collection—including keystroke monitoring—raises questions about employee privacy and consent. Meta has not yet disclosed which departments will be subject to tracking, the retention period for collected data, or whether participation is voluntary.
This development occurs as major tech companies intensify efforts to fuel large language model training. Meta joins other firms in exploring internal data sources as publicly available training material becomes increasingly contested due to copyright and ethical concerns.
The company has not released detailed guidelines on how employee data will be anonymized or protected.
Meta's child safety settlement is forcing social media competitors to adopt similar protections while shifting responsibility from users to tech companies for product design and age verification.
Nvidia is expanding its data center advantage beyond processor speed through smarter traffic management systems. The company's new generation hardware prioritizes efficiency over raw computational cycles.
A new social platform called Twitter has launched despite an ongoing trademark injunction case against Elon Musk's X. The service explicitly states it is not affiliated with X.
Nvidia CEO Jensen Huang interrupted an all-hands meeting to take a call from Donald Trump. Hours later, the president posted congratulations to Nvidia on Truth Social regarding the company's latest earnings.