Meta is using its own employees to train AI agents for ‘everyday tasks’

A new report claims that internal memos at Meta say the company will be harvesting data from employees to train artificial intelligence.

The training software Meta plans on using will go directly onto employees’ computers and will track what the employees are doing at work.

‘Agents primarily do the work and our role is to direct.’

The new directive will track U.S.-based employees’ activities on their computers, Meta reportedly told staffers, capturing mouse movements, clicks, and keystrokes. In turn, this data will train Meta’s AI models so that the automated agents can perform work tasks autonomously, Reuters reported.

In a statement to Return, a Meta spokesman said that if the company is “building agents to help people complete everyday tasks using computers, our models need real examples of how people actually use them.”

This includes not only the aforementioned mouse movements and clicks, but also “navigating dropdown menus,” for example.

The recent report stated that Meta will use a Model Capability Initiative that runs on work-related apps or websites and takes snapshots of what appears on the employees’ screens.

Meta described the initiative as launching an internal tool that will capture the mouse clicks and movements “on certain applications to help us train our models.”

At the same time, the spokesman said employee data would remain safe.

RELATED: West Virginia Republicans are betraying their voters for AI special interests

Artur Widak/NurPhoto/Getty Images

“There are safeguards in place to protect sensitive content, and the data is not used for any other purpose,” the spokesman asserted in his statement.

The data is only collected for “model training purposes” and will “not be used in performance reviews, and managers cannot access it,” the statement concluded.

Meta was asked to clarify what “everyday tasks” they were looking to have their AI agents perform and if this amounted to tasks that would otherwise be performed by a human, but the company did not provide an answer to those questions.

While the internal memos have not been published, Reuters claimed to have reviewed multiple, including one that was posted internally to the Meta SuperIntelligence Labs team.

“This is where all Meta employees can help our models get better simply by doing their daily work,” it allegedly said.

RELATED: Mother-daughter farmers reject eye-popping Big Tech bids: ‘I’ll stay … and feed a nation’

David Paul Morris/Bloomberg/Getty Images

Meta’s chief technology officer, Andrew Bosworth, allegedly shared a different memo this week that told employees internal data collection would increase at the company, as roles transform into directing AI agents to do work.

“The vision we are building towards is one where our agents primarily do the work and our role is to direct, review, and help them improve,” Bosworth reportedly stated.

Like Blaze News? Bypass the censors, sign up for our newsletters, and get stories like this direct to your inbox. Sign up here!