Meta says the Ray-Ban glasses that reportedly woke Meta AI by themselves are fixed
Meta says the issue behind reports of its Ray-Ban glasses starting Meta AI without the "Hey Meta" wake phrase has been resolved, and it has not said what the issue was. Its entire answer, given to Android Authority on October 13, was two sentences: "This issue has been resolved. People can manage, review and delete voice logs of their interactions with Meta AI via settings at any time."
Meta's own image of the Ray-Ban Meta glasses. Image: Meta
Those reports come from owners, with no number attached to them and no independent reproduction. Meta's statement does not describe the behaviour, does not say what caused it and does not say when a fix shipped.
Meta will not say what went wrong
There is no cause in the statement, no affected firmware or app version, no estimate of how many pairs were involved and no fix date. "Resolved" is Meta's own word, and because it carries no date there is nothing to tie it to a particular glasses or app update. The statement also names no model and no generation, so it does not establish which pairs were involved.
Meta's wording concedes little either. What it offers a route to deleting is "voice logs of their interactions with Meta AI", which describes a record of interactions the assistant had, and it stops short of saying anything was captured outside one.
Nor does the statement say what happens to audio from the activations the reports describe, which is the part an affected owner would most want settled. Meta's answer treats the question as a privacy-controls question rather than a fault report, and points at the delete button instead.
The voice log is the one thing an owner can act on
The instruction Meta does give is about the log: voice logs of Meta AI interactions can be managed, reviewed and deleted in settings at any time. Meta does not name the setting or say where to find it, so checking what a pair of glasses has sent is left to the owner. With no version number published to check against, there is nothing else for an owner to check.
Meta AI on the glasses is handled through Meta's phone app, which became the required companion for its glasses at version 292.1.0. The same app is where a separate setting decides whether people at Meta review photos sent to Meta AI, which is worth knowing before deciding how much of a log to keep.
The capture light would not have shown this
The glasses' most visible privacy control is the outward-facing capture LED, and Meta says that covering that LED disables the camera entirely. It governs the camera, not the microphones, so a bystander watching for it would have seen nothing during a voice interaction, and neither would the wearer.
A camera announces itself with a light that Meta now enforces, while a voice interaction leaves a trace only in a log on a phone.
Until Meta says more, that log is the only place an owner can see what their glasses have sent to Meta AI, and it is the only part of the answer that does anything.
