- Meta records employee clicks, keystrokes, and screen activity to train AI agents in real work behaviors
- The program is part of a broader push to build AI systems that can perform everyday tasks with minimal human input
- The move comes just ahead of announcements of layoffs in the company
Meta has started collecting everything its employees do while they go about their normal work to train its AI models, as first reported by Pakinomist. The Model Capability Initiative records mouse movements and clicks, keyboard keystrokes, and even occasional screenshots from computers used by Meta employees in the United States. The company wants to observe how people actually use software, then feed that behavior into AI models so they can learn to do the same things.
Essentially, Meta wants to make its systems more reliable for the small actions that define a workday. This means everything from navigating a menu and moving between windows to analyzing different website formats. These are not easily solved with text data alone.
“This is where all Meta employees can help our models improve just by doing their daily work,” the internal memo said.
The article continues below
Training your own successor
AI systems are moving from generating content to performing actions. They are being trained to perform tasks that have always required a person at a keyboard. It requires more examples than just a list of steps to complete a task. They must see how the work develops. Meta’s approach is to capture these steps directly, and turn everyday activity into training material.
Workplace monitoring has been around for a long time, but Meta’s approach is more detailed and more specific in its purpose. The system records the fine-grained interactions that are usually overlooked and builds a detailed picture of how tasks are performed in practice. According to the company, the data is not intended for performance evaluation, with safeguards in place to protect sensitive information.
The tracking program is part of a broader push at Meta to develop AI agents capable of handling everyday tasks. This Agent Transformation Accelerator focuses on building AI models for routine work across different tools and platforms.
The timing of the rollout is difficult to separate from other changes in the company. Meta is preparing to lay off around 10% of its global workforce, with more to follow. further cuts are expected later this year.
All-seeing AI eye
Beyond how Meta plans to use the data, the level of detail the program collects is unusually extensive. Logging every keystroke and mouse movement is more familiar to factories and warehouses than corporate offices. It’s a new level of visibility, and possibly an uncomfortably intrusive level for many.
That this is happening in the United States is not surprising. Companies here are generally only required to inform employees about the monitoring, whereas European labor and data protection rules place much stricter limits on this type of monitoring.
For Meta, the fact that examples of everyday tasks must be trained makes this monitor program the obvious choice. Employees may feel less comfortable having no choice but to subject every moment of their workday to observation and have that data used to potentially replace them and all of their colleagues.
If Meta’s program works as the company hopes, it is unlikely to remain unique to the company. Demand for real-world behavioral data will increase as AI capable of performing these tasks becomes more common.
Meta will create AI models that can fully mimic what human employees do at work. Whether that leads to more efficient tools or just a more insecure and potentially depressed workplace depends on how these AI models are implemented, but there’s no doubt they’ll see every click soon.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews and opinions in your feeds.
The best business laptops for all budgets
