# Meta Closes a Second Loophole That Let Ray-Ban Glasses Record in Secret > Source: > Published: 2026-08-28 23:11:55+00:00 *Meta's Ray-Ban smart glasses now stop recording when you cover the privacy light, but the harder question is whether a tiny blinking LED can carry this much trust.* Cover the light after you've already started filming, and until this week the camera just kept rolling. Not anymore. Alex Himel, Meta's vice president of augmented reality, said on Threads on August 27 that Meta is rolling out a fix that stops the camera the moment someone blocks the capture LED mid-recording. According to reporting from The Verge and 9to5Google, Himel said the glasses already refused to start recording if the LED was covered, but some users found the obvious workaround: start the video first, then cover the light. The new update closes that gap. The camera now stops working if the light is covered during a recording. That's useful. It's also late. Back in July, Meta said it was updating its AI glasses to disable the camera if the capture LED had been physically tampered with or destroyed, after modders and online tutorials showed people how to remove or defeat the warning light. Gizmodo reported in early August that cheap stickers could still make the indicator hard to see, while BGR had already covered the aftermarket modding problem in June. Meta says it removes ads, posts and Marketplace listings that promote tampering services, and can ban accounts that sell them. The point is plain enough: the LED is no longer just a design detail. It's the whole social contract. [China Forces Meta to Unwind Its 2 Billion Dollar Acquisition of Manus AI](https://startupfortune.com/china-forces-meta-to-unwind-its-2-billion-dollar-acquisition-of-manus-ai/) Beijing's NDRC ordered Meta to unwind its $2 billion acquisition of AI agent startup Manus, ruling that a Singapore relocation didn't shield the China-born company from Chinese jurisdiction. Manus confirmed on August 11 it's independent again, with affected user data set for deletion starting August 23. - [China forces Meta to unwind acquisition](https://startupfortune.com/china-forces-meta-to-unwind-its-2-billion-dollar-acquisition-of-manus-ai/) - [Meta's Manus AI deal unwound by](https://startupfortune.com/china-forces-meta-to-unwind-its-2-billion-dollar-acquisition-of-manus-ai/) France's privacy regulator, CNIL, made that argument in sharper language on June 29. Its public warning said smart glasses can capture images and voices of nearby people in an almost invisible way, because they look too much like ordinary glasses. CNIL also said technical signals such as a light have limited reach and can be insufficient. If you've ever tried to spot a blinking dot on a stranger's temple across a cafe table, you know the problem. Germany has pushed the issue from privacy warning into legal threat. HateAid, a German civil rights organization, said in August that it had filed criminal complaints against Meta, EssilorLuxottica and major German retailers over Ray-Ban Meta glasses, arguing that the devices could violate rules against disguised recording devices. German outlets including n-tv and ZDF reported the complaint, and ARD's Tagesschau separately reported that Hamburg's data protection commissioner, Thomas Fuchs, considers a full German ban possible if the glasses are treated as camouflaged cameras. Courts have been less patient than regulators. Bloomberg Law reported that New York's state court system banned smart glasses from its facilities starting July 20, covering any eyewear or headwear with recording technology. The Guardian later reported that courts in England and Wales had also banned Meta glasses, with HM Courts and Tribunals Service saying the devices would be confiscated on entry and returned when the wearer left. That's a blunt answer to a small light. ## Meta is trying to teach the room Meta is pairing the software fix with a public education campaign. The Verge reported that one Los Angeles billboard shows the glasses beside the line: The camera lets you capture the moment. The light lets everyone around you know when you do. It's an unusually defensive message for a product line that's selling fast. Bloomberg reported in February that Meta and EssilorLuxottica sold more than 7 million Ray-Ban and Oakley AI frames in 2025, with demand helped by new models introduced in the second half of the year. The Guardian put UK prices between £269 and £469. Those numbers explain why Meta keeps patching the problem rather than backing away from camera glasses. Too much is now riding on the frame. But you can't market your way out of consent. A phone has a camera, too, but you have to lift it, aim it - and somebody notices when you do. Glasses sit where ordinary glasses sit, which is exactly why they work as a hands-free camera and exactly why bystanders dislike them. The product's convenience and the privacy problem come from the same place. [Meta's Smart Glasses Trigger a Privacy Backlash It Can't Fix With Software](https://startupfortune.com/metas-smart-glasses-trigger-a-privacy-backlash-it-cant-fix-with-software/) Viral videos of strangers being secretly filmed have turned Meta's Ray-Ban smart glasses into what social media now calls "pervert glasses." A $60 mod let users disable the camera's warning light entirely, and Meta's July fix hasn't stopped senators, lawsuits, or a wave of UK venue bans. - [meta smart glasses privacy concerns](https://startupfortune.com/metas-smart-glasses-trigger-a-privacy-backlash-it-cant-fix-with-software/) - [recording people without consent laws](https://startupfortune.com/metas-smart-glasses-trigger-a-privacy-backlash-it-cant-fix-with-software/) Academic work is circling the same tension. A CHI 2026 paper by researchers from Tsinghua University and the University of Utah, titled Mind the Gap: Mapping Wearer-Bystander Privacy Tensions and Context-Adaptive Pathways for Camera Glasses, frames the issue as a conflict between the wearer's control and the bystander's awareness. That is the clean version of a messy public problem. The person wearing the glasses gets the interface. Everyone else gets a blinking dot. ## The next version could make this worse Here's the part Meta still has to answer. The Financial Times reported in July that Meta has been testing prototype smart glasses with a super-sensing feature that could take photos every few seconds and capture ambient audio so an AI assistant can answer questions about what the wearer has seen or heard. Android Authority and other outlets cited the FT's reporting that Meta executives had considered not activating the LED while that mode runs. Frankly, that's the tell. Meta can close every workaround users find in the hardware it already sells, and it should. But if future glasses collect ambient context for AI without giving people nearby a clear signal, the company hasn't solved the problem. It has moved it from abuse of the product into the product plan. Smart glasses are being pitched as the next interface for AI, the thing that sits closer to your eyes and ears than a phone ever could. That future only works if the people standing near the wearer trust the device on their face. Trust is the whole product. Right now, in Paris, Hamburg, New York and courts across England and Wales, that trust is being tested one doorway at a time. **Also read:** [Anthropic Says Claude Is Showing Early Signs of Self-Improvement](https://startupfortune.com/anthropic-says-claude-is-showing-early-signs-of-self-improvement/) • [Hackers Stole Data on 8.7 Million Manchester Airports Group Customers](https://startupfortune.com/hackers-stole-data-on-87-million-manchester-airports-group-customers/) • [Fed Chair Kevin Warsh's Inflation Warning Rattles Tech and AI Stocks](https://startupfortune.com/fed-chair-kevin-warshs-inflation-warning-rattles-tech-and-ai-stocks/)