Meta Is Prototyping ‘Super Sensing’ Glasses That Could Record Nonstop, Without Ever Lighting Up
Just hours after promising to make its AI glasses harder to abuse for secret recording, Meta is reportedly testing a very different prototype: AI glasses that continuously capture audio and snap photos every few seconds, without triggering any visible recording light.
Meta is developing prototype AI glasses that would keep recording almost continuously, according to a Financial Times report.
The glasses would collect audio nonstop while capturing photos every few seconds, letting a wearer later ask an AI assistant what they saw or heard, or use it to help reconstruct their day.
The concept is notable given the timing: just hours after Meta moved to make its recording LED harder to tamper with, reports suggest the company is considering turning that same white indicator off when its always-on recording features are active.
A Metadata Workaround, With Caveats
According to the Financial Times’ sourcing, one version of the system under discussion wouldn’t store the raw audio and video Meta collects, or even make it available to the wearer directly.
Supporters of this approach argue that this poses fewer privacy risks than storing full recordings, though Meta’s systems would still continuously process everything a wearer sees and hears to generate that metadata.
The report also says Meta is considering using data collected through the glasses to train its AI models, a question made more significant by the billions the company is spending to compete with OpenAI, Google, and Anthropic.
Notably, the feature could reportedly be enabled on Meta’s current glasses through a software update, meaning devices already in users’ hands could gain the capability without new hardware.
A Familiar Legal Gray Area
The prototype raises questions Meta hasn’t fully answered. Privacy experts cited in the report warn that an always-on wearable device like this could conflict with data privacy and biometric laws in some jurisdictions.
It also remains unclear whether Meta or the wearer would bear legal responsibility if the glasses violate wiretapping laws, especially since several U.S. states prohibit recording someone else’s conversation without their consent.
That uncertainty echoes concerns over Meta’s existing Ray-Ban glasses, which give bystanders no visual cue when the AI is analyzing its surroundings.
Meta argues privacy is protected by stripping out identifying details during processing, rather than limiting what the glasses capture in the first place.
It’s also not the first time Meta has explored features it later backed away from: Meta removed an inactive facial recognition system that had been quietly embedded in its Ray-Ban glasses platform after it faced backlash from the ACLU Coalition earlier this year.
A Pattern That’s Hard to Miss
The timing is what stands out most. The timing is what stands out most. The same week Meta introduced its LED safeguard, it also confirmed that Meta AI would use public Instagram photos and videos by default unless users opt out, as part of the rollout of its new Muse Image generator.
Together, the two announcements undercut the company’s message about protecting user privacy.
Whether the super-sensing prototype ever reaches consumers remains uncertain, and Meta’s plans could still change before anything ships.
But for a company already facing multiple state investigations and lawsuits over its AI glasses, a device designed to avoid signaling when it’s recording comes at an especially awkward time.
Source: Meta tests ‘super sensing’ AI glasses that can capture every moment
