Tech1 hr ago

Meta to Monitor Employees' Keystrokes for AI Training Amid $140bn AI Spend Push

Meta announces plans to monitor employee keystrokes and mouse clicks to train its AI models, amid a planned $140 billion investment in AI by 2026.

Alex Mercer/3 min/GB

Senior Tech Correspondent

TweetLinkedIn
Woman with dark curly hair, wearing tan polo neck jumper, works on a computer in an office

Woman with dark curly hair, wearing tan polo neck jumper, works on a computer in an office

Source: BbcOriginal source

Meta will track employee keystrokes and mouse clicks on company devices to train its artificial intelligence models, aligning with its substantial $140 billion AI investment push.

Meta, the parent company of Facebook and Instagram, is implementing new internal monitoring for its employees in the UK and globally. This program records activity such as keystrokes and mouse movements on company computers and internal applications. The company confirms this collected data will exclusively train its artificial intelligence models, a move designed to enhance AI capabilities by providing real-world usage examples.

This initiative coincides with a substantial financial commitment to AI development. Meta intends to spend approximately $140 billion on AI in 2026. This investment figure represents nearly double its AI expenditure from the previous year, highlighting a focused, aggressive shift towards AI advancement. A Meta spokesperson clarified the logged data's sole purpose is AI training, with specific safeguards in place to protect sensitive content from other uses.

This new tracking has generated internal reactions. One unnamed Meta employee described the monitoring of their keystrokes for AI training as feeling "very dystopian." This sentiment emerges as employees anticipate further job reductions across the company, adding a layer of concern regarding workplace conditions. The program, known internally as the Model Capability Initiative (MCI), aims to provide Meta's developing AI models with practical, real-world examples of how people interact with computer systems. This data is considered essential for building effective AI agents capable of assisting with diverse tasks.

Meta's leadership has consistently positioned AI as a central pillar of future growth and operational efficiency. CEO Mark Zuckerberg previously indicated that AI could dramatically alter work processes, potentially enabling smaller teams to achieve what once required larger groups. The implementation of employee monitoring underscores the company's aggressive strategy to gather proprietary data for its advanced model development. The long-term impact on workplace culture, employee trust, and the tangible improvements in Meta's AI products will be critical areas to observe as the company continues its rapid AI expansion.

TweetLinkedIn

More in this thread

Reader notes

Loading comments...