Meta’s employee-tracking AI tool put on hold after exposing sensitive data internally

Meta has paused a controversial employee monitoring program after it discovered the program left potentially sensitive information, collected by employee hardware, accessible to all Meta employees.

Meta’s Model Capability Initiative (MCI) left transcriptions, employee conversations, and performance data accessible inside the company, first reported by Business Insider. And, according to Quartz, one employee’s personal tax and medical records.,

“We have carefully designed this program with privacy safeguards and while we have no indication at this time that any data was improperly accessed by Meta employees, we’re pausing it while we investigate,” Tracy Clayton, a Meta spokesperson, told HR Brew via email.

The discovery represented an unfortunate confirmation of some Meta employees’ fears expressed in the wake of the program’s launch in April. Meta announced the initiative to support development of Meta’s AI modeling, collecting employee keystrokes, mouse activities and clicks, and even occasional screenshots as training data.

The program, which was mandatory for most staff, sparked a backlash from some employees who raised privacy concerns. Morning Brew previously reported Meta CTO Andrew Bosworth told employees the company behind Facebook, Instagram, and WhatsApp was ramping up its internal data collection in order to prepare for the future where “agents primarily do the work” and Meta human employees “direct, review and help them improve.”

The pause comes as Meta deals with the fallout from a contentious and challenging broader AI initiative roll out. Tech Brew reported earlier this month that Bosworth copped in an internal memo to the company’s “atrocious job” rolling out its AI reorganization, which involved layoff announcements, surprise reassignments, and the new surveillance technology that ostensibly trained employees’ technological replacements.

Tech companies aren’t the only ventures dealing with the real workforce challenges of rolling out new AI initiatives and programs. HR Brew previously reported some execs are looking to shakeup leadership in their HR departments to specifically address the human issues that arise amid the AI transformation.

“Something is coming, or things are going to be new, and there’s no blueprint on how to get there. It’s unclear what work is going to look like in one or two years. Many of the eyes in the room are looking at [CHROs],” Ted Moore, managing director and global lead of the HR practice at RRA, previously told HR Brew.