Microsoft OneDrive's AI Face Recognition Feature Raises Privacy Concerns Over Opt-Out Limits

Microsoft’s latest experiment inside OneDrive has users questioning just how much control they really have over their data. The company is quietly testing an AI-driven face recognition feature that automatically groups photos by people. However, a detail buried in its settings has triggered alarm among users.

According to a report from Slashdot, the preview users found that the face recognition setting can only be turned off three times per year. That means anyone hoping to frequently change their privacy preferences might hit a limit. Needless to say, that’s an unusual restriction for a feature still in testing.

The discovery quickly raised eyebrows among privacy advocates. Speaking to Slashdot, EFF’s Thorin Klosowski said features like this should “always be opt-in,” adding that “companies should provide clear documentation so users can understand the risks and benefits to make that choice for themselves.”

Klosowski also criticized the toggle restriction, saying people “should be able to change those settings at will” as privacy needs can shift over time.

Following up on the story, Windows Central came across Microsoft’s own documentation that talks slightly different. The company says users can disable face recognition freely via the OneDrive settings page, and that doing so permanently deletes all face data within 30 days.

Microsoft further insists that the feature doesn’t train any shared AI models, all data stays local to your account. For now, the AI face grouping tool remains in limited preview, and not everyone will see it yet. But as Microsoft prepares for a wider rollout, users are already asking for clearer privacy controls.

via: PC Gamer