Meta is tightening the privacy controls on its AI-powered smart glasses again, this time targeting a workaround that let wearers keep filming people without their knowledge. According to an announcement from Alex Himel, Meta's vice president overseeing the glasses line, a new update will cut off the camera entirely if a user covers the device's white capture LED while a recording is already in progress.
Why the Fix Was Needed
Meta's smart glasses have always included a small LED that lights up on the front of the frames whenever the camera is capturing photo or video, giving bystanders a visual cue that they might be on camera. Until now, that safeguard had a gap: the glasses would refuse to start a new recording if the LED was already covered, but a wearer who began filming first and then discreetly placed a finger or piece of tape over the light could keep the camera running unnoticed. That gap turned a safety feature meant to protect the public into something that could be defeated with a single motion.
The consequences of that gap became a public relations problem for Meta. As the glasses grew more popular, so did reports of people using them to record strangers without consent, film unsuspecting members of the public for pranks, or harass people in everyday settings. The backlash was loud enough that the glasses picked up unflattering nicknames online, and pressure mounted on Meta to prove its "privacy indicator" wasn't just decorative.
What's Actually Changing
With the new update, covering the LED at any point during a recording — not just before one starts — will immediately stop the camera. Himel described it as a direct response to the small number of users who were "going to these lengths" to bypass the indicator, framing their persistence as evidence that the safeguard was working as intended rather than a sign it needed strengthening. Either way, Meta is now closing the specific technical gap that made the bypass possible.
This builds on a change Meta rolled out in July, which disabled the glasses' cameras altogether if the device detected physical tampering with the LED itself, such as scratching, painting over, or otherwise damaging the light. Combined, the two updates form a broader push to make the capture indicator harder to defeat by any means, whether digital or physical.
An Awareness Campaign, Too
Alongside the software fix, Meta has launched what it's calling a Capture LED awareness campaign, aimed at educating both wearers and the general public about what the light means. New ads posted to Meta's own channels use the tagline "no light, no capture," part of an effort to reframe the glasses as a product "designed to be noticed" rather than designed for covert surveillance. The campaign is as much about public trust as it is about the underlying technology — a tacit acknowledgment that a blinking LED only works as a privacy safeguard if people actually understand what it's telling them.
The Bigger Picture
Meta's update lands at a moment when AI-powered wearables are being positioned across the industry as a likely successor to the smartphone, precisely because they can constantly see and hear their surroundings. That capability is also what makes them uniquely sensitive from a privacy standpoint: unlike a phone, which usually has to be visibly pointed at someone, glasses can capture video without an obvious gesture at all.
For Meta, closing this particular loophole is a relatively small technical fix. The larger challenge — convincing not just wearers but everyone around them that an always-on camera on someone's face can be trusted — is likely to keep testing the company as smart glasses become more common, and as people keep looking for new ways around whatever safeguards come next.
