Meta Plans to Record Employee Keystrokes and Mouse Movements for AI Training
Meta turns inward for AI training data
Meta is planning to collect a new kind of training data for its AI models: information from its own employees’ mouse movements and keystrokes. The company wants to use those inputs to help build more capable and efficient AI systems, according to reporting first published by Reuters.
The move underscores how aggressively tech companies are searching for fresh sources of training data, which remains the fuel for modern AI systems. In Meta’s case, the company says it wants real examples of how people use computers when building agents designed to help complete everyday tasks.
When asked about the plan, a Meta spokesperson told TechCrunch that the company is launching an internal tool that will capture inputs on certain applications, including mouse movements, button clicks, and navigation through dropdown menus. The spokesperson said safeguards are in place to protect sensitive content and that the data will not be used for any other purpose.
Even with those assurances, the plan raises obvious privacy concerns. Recording how employees interact with their computers can reveal far more than simple workflow patterns, especially when the data comes from tools people use throughout the workday. The proposal adds to a growing debate over how far companies can go in collecting behavioral data in the name of AI development.
The report also fits into a broader pattern across the industry. Last week, it was reported that older startups are being mined for corporate communications such as Slack archives and Jira tickets, with those materials then converted into AI training data. Meta’s approach takes that same impulse one step further by turning routine employee activity into a training set.
For Meta, the bet is that observing how people actually use computers will help its models better understand task completion in real-world settings. For critics, the question is whether the race for better AI is pushing companies to normalize surveillance practices that would once have seemed off limits.
Sources:
Browse privately with Doppler VPN — no logs, one tap connect.