Meta investigates employee data concerns over mouse-tracking AI initiative

Why was Meta tracking employee mouse movements?

What if your everyday computer clicks were helping train artificial intelligence? That was the idea behind an internal Meta program that recorded employee activity as they worked. The initiative was designed to teach AI how people use computers in real situations.

But the experiment suddenly became controversial after an employee raised concerns that sensitive information may have been exposed. Meta has now paused the program and launched an investigation into what exactly happened behind the scenes.

The program was meant to make AI act more human

The internal effort was called the Model Capability Initiative, or MCI. It reportedly began in April and aimed to improve AI systems in tasks that humans do naturally while using a computer.

Meta wanted its AI to understand actions such as selecting items from a dropdown menu, moving around software interfaces, and using keyboard shortcuts. Employees were effectively helping train the systems simply by doing their normal jobs.

A worker claimed sensitive information was exposed

The concerns escalated when an employee reportedly filed a high-priority security incident report. The worker claimed the program had access to far more information than employees were originally told.

The report allegedly mentioned prompts, transcriptions, private conversations, and performance information. It also claimed that some of the collected data was stored in an unencrypted form, raising serious questions about privacy protections.

Claims even included tax and medical information

One of the most alarming parts of the report involved personal information. The employee said tax and medical details had been accessed through work computers and could have been included in the data being collected.

The worker claimed employees had been assured that aggressive filtering would protect private information and limit data use to legitimate business purposes. Those assurances are now under fresh scrutiny.

Meta says it found no evidence of misuse

Meta says it has no indication that employees improperly accessed the information collected through the initiative. The company stressed that privacy safeguards were built into the program from the beginning.

Even so, the company decided to pause the project while it investigates the claims. The move suggests Meta believes the concerns are serious enough to warrant a closer review of how the system operated.

The pause did not immediately stop the program

Interestingly, pausing the initiative did not instantly shut it down for everyone. Meta acknowledged that it would take time to stop the program across its systems.

Reports indicated that some employees could still see the software running even after the investigation was announced. That detail added another layer of concern about the company’s ability to quickly disable the system.

Employee monitoring tools are becoming more common

Many companies already use software that measures productivity or tracks digital activity. Some tools can record keyboard activity, application use, and even screenshots of employee devices.

Meta’s situation highlights how these technologies can become controversial when workers believe data collection goes beyond what they expected or includes sensitive information.

Little-known fact: A 2025 survey by the American Psychological Association found that more than half of monitored workers worry about how employers use collected data.

The case raises difficult questions about AI development

Building better AI often requires huge amounts of real-world information. Yet companies also have a responsibility to explain exactly what data is being collected and how it is being protected.

The concerns surrounding MCI show that even internal experiments can trigger significant privacy debates when employees fear their personal information may be involved.

Meta is now under pressure to provide answers

Questions remain about exactly what information was collected, how it was stored, and whether any safeguards failed. The company has not said how long its investigation may take.

Employees and industry observers will likely be watching closely for details because the outcome could shape how future AI training programs are designed inside major technology companies.

The controversy could influence future workplace policies

The incident arrives at a time when businesses are increasingly experimenting with AI and data collection. Workers are also becoming more aware of how much information employers can potentially gather.

If companies want employees to participate in AI projects, transparency and strong privacy protections may become just as important as the technology itself.

Questions about how companies manage AI are growing louder. Take a look at why OpenAI’s safety failures are under fresh scrutiny, for more on the debate unfolding behind the scenes.

A reminder that AI progress can create unexpected risks

Meta’s paused initiative shows how quickly an effort meant to improve artificial intelligence can turn into a debate over privacy and trust. The company says it has found no evidence of improper access so far.

The investigation is still unfolding, but it already serves as a reminder that gathering data for smarter AI systems can create unexpected concerns.

Worried about your privacy? Here are 8 things you should never share with an AI chatbot if you value privacy.

What do you think about Meta’s employee tracking controversy? Share your thoughts.

This slideshow was made with AI assistance and human editing.

Don’t forget to follow us for more exclusive content right here on MSN.

Read More From This Brand:

• Meta explores unconventional data center construction for AI growth
• Meta’s smart glasses draw privacy concerns after code discovery
• Meta job cuts spread across multiple divisions during restructuring