Meta has confirmed and fixed a bug that caused its Ray-Ban smart glasses to record users' conversations without the owner ever saying "Hey Meta" — but new complaints suggest at least some wearers believe the problem is not entirely behind them.
The company acknowledged the issue after users documented the glasses capturing audio and generating transcripts of ordinary conversations, despite the feature that allows recording without the wake word being switched off. For more context on this story, see our ongoing latest AI developments.
What users reported
The reports began on Reddit, where one user described Meta's smart glasses recording and transcribing conversations that were never directed at the assistant. According to the post, as reported by Mashable, the setting that permits recording without saying "Hey Meta" was turned off at the time, and the glasses captured the audio anyway.
More troubling for the wearer, the recordings were not discarded. The user said full audio recordings and text transcripts of the conversations were saved to their Meta AI conversation history, and that the glasses' white activation light turned on repeatedly during normal conversation with a friend.
"I have observed the same behavior on THREE different pairs of Ray-Ban Meta glasses, including 2nd- and 3rd-generation models," the user wrote, in a post cited by multiple outlets. The issue therefore did not look like a single defective unit but a software problem affecting multiple hardware generations.
Other users in the comments reported similar experiences, saying they assumed they were the only ones affected until the thread made clear the pattern was broader.
Meta: 'This issue has been resolved'
A Meta spokesperson confirmed the glitch, telling Android Authority: "This issue has been resolved. People can manage, review and delete voice logs of their interactions with Meta AI via settings at any time."
The company has not explained what caused the glasses to capture audio when they should not have, how long the flaw existed, or how many devices were affected. That silence has frustrated privacy advocates, who note that a microphone activating on its own is among the most serious failure modes a wearable camera-and-audio device can have.
'Something weird is going on'
The story took a new turn over the weekend. According to a New York Post report published October 17, wearers in three separate Reddit threads claimed that multiple generations of Ray-Ban Meta glasses continued to record conversations without being activated, with some attributing the behavior to a recent firmware update.
"Mine randomly activates the internal light now without any input from me," one user wrote. "It does not make the wake-up noise voice command would. The internal light just turns on. Something weird is going on!"
Another said they had reset the glasses only to see the behavior persist, reporting that the LED would light up in situations where "Hey Meta" was never spoken. One wearer told the Post they were preparing to file a class-action lawsuit if it could be shown the issue was not isolated.
Meta repeated the same statement to the Post, saying the issue had been resolved and pointing users to the settings where voice logs can be managed, reviewed, and deleted. The company did not address the sleep-mode claims specifically or the firmware theory.
Why the indicator light matters
Ray-Ban Meta glasses are designed around a simple privacy contract: a white LED illuminates whenever the device is capturing, giving people nearby a visual cue that recording is in progress. Privacy researchers have criticized that design for years, noting how easy it is for a capture indicator to be missed in daylight or obscured by hair or a sticker.
The reported bug breaks the contract differently — and arguably worse. In these accounts, the light did illuminate, but only because the microphone had activated on its own during private conversations. The user's own device became the surveillance vector, with transcripts preserved by default rather than discarded.
A familiar pattern for Meta wearables
The episode is not the first privacy controversy for Meta's glasses line. Earlier this year, an investigation revealed that videos recorded by Meta smart glasses — including intimate footage and clips filmed in bathrooms — had been sent to offshore contract workers in Kenya for AI data labeling, meaning third parties had viewed users' most private moments without their knowledge.
That history colors how the current reports are being received. Smart glasses already occupy an uncomfortable privacy position, since they look like ordinary eyewear while carrying a camera and microphone. A device that records even its own owner without prompting undermines the core assurance Meta offers both wearers and bystanders.
What wearers can do now
Meta says voice recordings and transcripts can be reviewed and deleted at any time through the settings associated with Meta AI, and wearers who are concerned about unintentional captures should check their conversation history for entries they did not create deliberately.
Beyond that, the burden falls back on Meta to explain what failed. A confirmed bug that recorded audio without a wake word, followed by fresh claims of recording in sleep mode, is the kind of pattern that usually warrants a public post-mortem. Until the company offers one, the people most likely to trust smart glasses enough to wear them every day will keep watching that little white light — and wondering what it is doing on.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
