Meta's smart glasses were recording conversations without the wake phrase, company admits

Meta's smart glasses have been revealed to be recording conversations without users' authorization, the company admits. The issue was discovered when a user noticed their voice was captured by their Meta Ray-Ban glasses, despite the "Hey Meta" wake phrase not being used. The recorded conversations appeared in the Meta AI app as audio recordings and text transcripts.

The issue persisted even after the "Respond without Hey Meta" setting was disabled, and the device's white notification light was activated randomly during normal speaking. This problem was found across three different pairs of smart glasses, including older and newer models. The issue seems to have arisen following a firmware update.

When Meta support was contacted, they advised users to disable microphone and camera permissions, turn off Wear Detection, delete voice history, and disable automatic updates. However, this response was criticized as it would disable the product's core functionality and fail to address the problem. Several users confirmed experiencing the same issue with their Meta Ray-Ban glasses.

Meta has since stated that the issue has been resolved, but they did not provide further details on the cause. The company advised users to manage, review, and delete voice logs of their interactions with Meta AI through settings. The use of smart glasses in public is facing increasing scrutiny, with Norway considering temporary bans in specific areas, and courts banning them across the US.

Meta has faced previous privacy concerns with its smart glasses, including the case of contractors in Nairobi who were paid to study raw footage captured through Meta's "live AI" feature.