Meta's guidance treats the capture LED as a condition of using the camera at all: covering, damaging, or otherwise tampering with the light disables the camera automatically, according to Meta's corporate blog. The light itself is a blinking white LED with no off switch, built to signal to bystanders that recording is underway.

Losing camera access doesn't kill the rest of the hardware. Audio, the AI assistant, and Bluetooth headphone functions all keep working, so a bricked pair becomes what PetaPixel called "essentially an expensive pair of Bluetooth headphones," per PetaPixel. In short, the Ray-Ban smart glasses camera gets bricked; nothing else does.

Tampering methods have escalated. Early workarounds involved a strip of tape over the LED, which Meta already blocked; some users then moved on to drilling out the LED hardware entirely, The Next Web reported. That progression from tape to hardware modification is what pushed Meta toward continuous monitoring rather than a single check, according to VR.org's technical analysis.

Meta's first fix, shipped in July, made an intact LED a precondition for recording. If the light was covered or destroyed when someone tried to film, the camera simply wouldn't start, per Meta's corporate blog.

Users found the obvious gap almost immediately: start recording, then cover the light. Himel addressed it directly in a Threads post, writing that "the camera won't start recording if the LED is blocked, but some people try to bypass this by starting a recording and then covering it up," per PetaPixel and Medianama.

An update that began rolling out in late August closed that gap. VR.org's analysis describes the shift as promoting the LED check from a one-time gate into a continuous requirement that has to hold for the entire session, meaning something on the glasses is now sampling the light's status on a loop with the authority to cut a recording off mid-stream, according to VR.org. Meta hasn't published the sampling interval, detection method, or firmware version behind the change. VR.org, citing Engadget, reported that a two-dollar sticker defeated the light without triggering detection, a considerably shorter list of steps than drilling out a component.

Himel isn't claiming victory. "I don't think we'll ever be done... There's always new ways to get around the safeguards that are in place," he told Semafor, per PetaPixel. Elsewhere he called the dynamic "a bit of a cat-and-mouse game," per VR.org.

A Reddit post illustrates the continuing dispute. Someone who claimed to film people without their consent complained about being "blown" by the update four days ago and asked whether any workarounds remained. Commenters called him creepy, though a few suggested recording without the light is still possible, per PetaPixel.

Firmware enforcement addresses tampering. It does nothing for recordings made without any tampering at all, which is the category behind the sharpest legal pressure Meta is facing right now.

On August 24, the Internet Freedom Foundation sent Meta a legal notice on behalf of a man identified as Mr. X, over footage filmed at a café in Delhi's Khan Market on July 24, Medianama reported. The notice names both Meta and the creator who allegedly recorded him. According to the notice, Mr. X had no idea the Ray-Ban Meta glasses were filming the interaction, and nothing in the notice alleges the LED was tampered with.

The footage went up on Instagram two days after it was recorded and, per the notice, crossed 419,000 views, 15,700 likes, and more than 500 comments. It seeks a takedown, disclosure, an apology, preservation of evidence, and roughly Rs 2.05 crore in damages.

The legal theory is untested. The notice argues Meta could be liable as the maker and seller of the physical device, separate from its role hosting the video on Instagram, falling outside the protection of Section 79 of India's IT Act, according to Medianama. No Indian court has yet ruled on whether a smart-glasses maker can be held liable on that basis.

This isn't the first such incident. In March, someone secretly recorded a person identified as Shubnam during a transgender rights protest using Ray-Ban Meta glasses, an episode that raised the same question the Delhi notice now poses formally: who bears responsibility when one company controls both the camera and the platform distributing the footage, per Medianama. Meta has said users are responsible for using the glasses safely and respecting others' privacy.

Consent isn't the only concern. A BBC report from March, cited by Medianama, said the U.K. data watchdog contacted Meta after reports that outsourced contractors in Kenya could view sensitive footage, including intimate material, captured through the glasses. Meta acknowledged that subcontractors sometimes review images and video to improve the product.

Institutions aren't waiting for a court ruling before drawing their own lines. New York's court system barred camera eyewear from more than 1,240 courthouses in July, and Philadelphia's courts followed, VR.org reported. Last week, a New York nightclub threatened to permanently ban anyone trying to enter wearing the glasses, and a budget pub chain in the U.K. has banned them outright, according to PetaPixel.

Meta's response extends beyond firmware. The company says it has removed thousands of posts linked to harassment and pranks, disabled the accounts involved, and blocked related search terms and hashtags, per Medianama. It also says it removes ads and Marketplace listings advertising LED-tampering services and takes legal action against people or businesses selling tools built to defeat the capture light, on and off its own platforms, according to Meta's corporate blog.

Himel frames the low number of confirmed tamperers as evidence the approach works: "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," per VR.org. That's his argument, not a settled outcome, given that Meta is also defending a class action over contractor review of footage and European regulators have their own inquiry open, per VR.org.

The updates also leave a broader design question unresolved: a working camera on someone's face, disclosed to strangers by nothing more than a light that depends on Meta staying ahead of whoever tries to defeat it next. Rival hardware is testing that question from the other direction. Even Realities, a Shenzhen startup founded by former Apple engineers, raised $150 million at a $1 billion valuation in a round led by Meituan and Tencent, The Next Web reported. Its glasses skip the camera entirely, projecting text and graphics for navigation, translation, and notifications, eliminating camera-based recording risk by omitting the lens. The trade-off is real: TechRadar's reviewer judged the display-only design not yet worth wearing, finding it more limited and less polished than camera-equipped models, per the same report.

Apple has floated a different way to make a capture signal harder to spoof. A patent application published on August 13 describes generating a capture cue inside a dedicated secure audio processor, then verifying through a separate secure microphone, over what the filing calls an untamperable path, that the cue actually played, according to VR.org. It's a filing, not a shipping product, but it describes an alternative to a single visible LED that a screwdriver or a two-dollar sticker can try to defeat.

Two firmware updates in seven weeks is a fast cadence for policing one light, per VR.org. None of it touches what the Delhi notice describes: no alleged tampering, just a camera doing exactly what Meta's design allows it to do, and a bystander who never saw it coming. Courts, venues, and regulators have already started drawing their own lines rather than waiting on Meta's next patch. Whether face-worn cameras become something the public tolerates now rests less on engineering than on what those institutions decide next.