Meta Plans to Monitor Employee Keyboard Activity and Mouse Movements for AI Development Amidst Staff Resistance
Meta Platforms Introduces Controversial AI Training Tool for US Employees
Meta Platforms has rolled out a new software on its US employees' computers that tracks keystrokes, mouse movements, and screen activity to train its AI systems. The move has sparked widespread backlash within the company, according to internal communications obtained by Business Insider.
The programme is designed to help AI models better understand how people interact with computers, including using keyboard shortcuts and choosing from dropdown menus. The internal announcement stated that the programme is necessary to train agents to understand how people complete everyday tasks using computers. Employees can help the company's models get better simply by doing their daily work.
The tool, which started rolling out on April 21, captures computer inputs like mouse movements, click locations, and keystrokes, as well as screen content for context. It only applies to US-based full-time employees (FTEs) and contingent workers, and not to employees' phones.
Read also: South Korean AI Startup Posts Near-Perfect Score on JEE Advanced 2025 Exam Paper
However, employees have voiced concerns and unease about the initiative. The top-rated comment in response to the internal announcement was "This makes me super uncomfortable. How do we opt out?" According to Business Insider, the "angry-face" emoji was the most common reaction to the original announcement.
Meta CTO Andrew Bosworth responded to the concerns, stating that there is no option to opt out of the programme on work-provided laptops. This comment received a mix of negative reactions from employees, indicating that they are not happy about the initiative.
Despite the backlash, Meta has assured employees that the collected data will not be used for any other purposes. The company has established safeguards to protect sensitive content and has informed employees of this upon onboarding. The programme is viewed as an extension of existing policies rather than a brand-new policy shift.
The software is limited to a list of commonly used work applications, such as Gmail, GChat, and Metamate, an AI assistant for employees. It only applies to computers, not to employees' phones. Employees can learn more about the tool and its privacy safeguards by checking out the company wiki and FAQs.
Read also: AI Visionaries Support Startup Developing Event Prediction Models
| Comparison of Meta's AI Training Tool | Original Announcement | Current Status |
|---|---|---|
| Purpose | To train AI models on real examples of human computer interaction | To help AI models better understand how people interact with computers |
| Data Collection | Captures computer inputs like mouse movements, click locations, and keystrokes, as well as screen content for context | Same as above |
| Applicability | US-based FTEs and contingent workers | Same as above |
| Phone Monitoring | No | No |
Note: The comparison table highlights the key differences between the original announcement and the current status of Meta's AI training tool.
Investor Takeaway
Investors should be cautious of potential employee backlash and its impact on Meta's operations.
More in General
South Korean AI Startup Posts Near-Perfect Score on JEE Advanced 2025 Exam Paper
AI Visionaries Support Startup Developing Event Prediction Models
Bengaluru-based QNu Labs Pursues Large-Scale Quantum Network Deployment for Military Applications
