Meta now says its AI-powered smart glasses will shut off the camera when someone tampers with the recording indicator light, a fix that came only after reports that users were filming strangers and posting the footage online.
The company rolled out the update after CBS News reported that a woman discovered a video of herself on social media, filmed without her knowledge by a man wearing Meta's AI glasses. She told CBS News she had no idea she was being recorded and never gave permission.
Her reaction was three words long: "I had no say." That single sentence captures a privacy problem Meta has spent months trying to patch over with software tweaks and public relations, while continuing to sell a product that puts a camera on someone's face and points it at everyone else.
Meta's glasses include a small LED light designed to signal bystanders when the wearer is recording. The idea is simple: if the light is on, people nearby know they might be on camera. But users figured out they could cover or block the light, and record without anyone noticing.
Meta says it recently updated the glasses so the camera deactivates whenever tampering with the LED is detected during or after recording begins. Before the update, the system only locked the recording function if the light was blocked before filming started. Once recording was underway, covering the light did nothing to stop it.
CBS News senior crime and public safety coordinating producer Anna Schecter tested an earlier version of the feature and confirmed the gap. When she covered the light after recording had already begun, the camera kept rolling. Meta's update is supposed to close that loophole.
Alex Himel, Meta's vice president of wearables, posted on Threads about the changes:
"We're going to keep rolling out updates like this to make sure the capture LED can reliably alert bystanders when photos or videos are being captured for someone's gallery."
Himel also downplayed the scope of the problem, saying only "a tiny minority of people" try to circumvent the protections. He framed the workarounds as proof the system works, arguing that the effort required to defeat it shows Meta's approach is sound.
"Only a tiny minority of people try to get around these protections, and the fact they have to go to these lengths shows our efforts are working."
That is a convenient way to describe a product whose privacy safeguard failed a basic test by a news producer.
Meta told CBS News by email that the updated camera-deactivation feature is already producing results. The company said businesses that had been offering LED tampering services, helping users disable or obscure the recording light, have shut down because the camera now deactivates when tampering is detected.
"These investments are already having a deterrent effect, companies that were offering LED tampering services have shut down because the camera now deactivates when tampering is detected."
Meta did not name any of those companies, and CBS News did not independently verify the claim. The company also said it removed thousands of harassing posts tied to pickup artist and prank content, and blocked search terms and hashtags, including those associated with "pickup artist" and "rizz" posts, linked to violating content. Meta also launched an education and awareness campaign about the LED feature.
None of those steps address the core issue: the glasses put a discreet, always-available camera on someone's face in a form factor that looks like ordinary eyewear. A tiny LED light, even one that works perfectly, is not a meaningful substitute for consent. Most people do not know what the light means, and in a crowded room or on a busy sidewalk, they are unlikely to notice it at all.
Meta built a wearable camera and shipped it to consumers with a privacy model that depends entirely on a small light and the good faith of the person wearing it. When users defeated the light, Meta patched the software. When the patch failed a journalist's test, Meta patched it again.
The woman who found herself filmed and posted online without her knowledge did not fail to notice a tiny LED. She was recorded by someone who chose to ignore whatever safeguards existed. No software update changes the incentive structure that made that possible in the first place.
Government action on the matter is absent from the record. No laws invoked, no charges filed, no regulatory response. Meta is policing its own product on its own terms, deciding what counts as misuse and announcing its own enforcement results. That is not accountability, it is a press release.
When the company that profits from the product is also the only entity deciding whether the product is safe, the public is left with exactly what that unnamed woman got: no say.
