Meta has recently told employees in the United States that it may begin tracking their keystrokes to help train future artificial intelligence tools. The company says this data could show how people work and improve its systems. However, a new report suggests the idea has raised concerns among staff and is meeting growing resistance. Many employees feel uneasy about being closely monitored during their daily tasks. They also worry that patterns from their own work could be used to build systems that might later replace parts of their roles. The current situation has created a rising tension between company goals and employee trust, while management defends the plan, stating that it’s necessary for innovation.
As reported earlier today, the tool is known as ‘Model Capability’ and is used to record keystrokes, mouse movements, clicks, and screen activity of the employees as they carry out their daily tasks. Moreover, this data will later be used to train AI models to better mimic human behaviour on computers. The company says that the idea is simple: if AI can observe how people complete real tasks, it can learn to do them more efficiently.
Also read: OpenAI CEO Sam Altman takes dig at Anthropic Mythos AI, calls it fear-based marketing
While the company is aiming to further expand its AI push and grab a huge leap in the AI sector, the rollout has not been welcomed by everyone. Internal discussions show that several employees are uncomfortable with the level of monitoring. One of the most common concerns raised by the employees was whether there is any option to opt out of this programme.
However, the company leadership has confirmed that there is no such choice for work-issued devices and the keylogger will be installed for everyone in the company irrespective of their designation. This has further added to the unease of the employees as reactions from them included frustration and concern about privacy.
Meta, on the other hand, defended the initiative by saying that similar monitoring has existed before and that this is only an extension of current practices. The company also said there are safeguards in place and that the collected data will only be used to improve AI systems and will not be used for other purposes, silently hinting about the evaluation.
Also read: Anthropic investigates alleged unauthorised access to its Mythos AI model: Here is what happened
It’s also said that the tracking tool is limited in scope, as it works only within certain approved applications like email, chat tools, and coding platforms. Furthermore, it does not track activity on personal phones. Still, for many employees, the issue goes beyond where tracking happens.
There is a bigger worry that these efforts could make automation happen faster at work, as by helping AI learn quickly, employees feel they might be digging up their own grave. Moreover, it also points to a larger problem in the tech industry, where improving AI often brings concerns about how it will affect workers.