By 
Meta has confirmed that a mandatory software update has permanently disabled the cameras on thousands of its AI-powered glasses, after the company detected users tampering with the small recording indicator light built into the devices. The confirmation came on Tuesday, with Meta stating that the enforcement action affected less than one tenth of a percent of all Meta AI glasses sold globally.
The figure sounds small in percentage terms, but the absolute numbers are not trivial. Meta has not disclosed how many units it has sold in total since the glasses launched in 2021, though the company has said it sold seven million pairs in 2025 alone. Applied to that base figure, a tenth of a percent works out to roughly 7,000 devices with cameras that no longer function, a number Semafor first reported.
What triggered the shutdown
Every Meta AI glasses model built since the second-generation Ray-Ban Meta line launched in 2024 has shipped with a white capture LED next to the lens. The light is designed to switch on automatically whenever the camera records a photo or video, giving bystanders a visual cue that they are being filmed. It has been Meta's primary answer to years of criticism that its glasses make it too easy to record people without their knowledge or consent.
That criticism has not gone away. Users found ways to defeat the light almost as fast as Meta could reinforce it. Some drilled the LED out of the hardware entirely. Others simply placed a sticker or piece of tape over it. Meta's first line of defense against this was a software check that disabled the camera if it detected the light was obstructed at the moment recording began. That check had a hole in it: people learned they could start recording with the light exposed, wait a few seconds for the software to register the light as functional, then cover it and keep filming.
The update Meta confirmed this week closes that specific loophole by continuing to monitor the LED throughout a recording session rather than only checking at the start. If the light is blocked or disabled partway through, the camera now shuts off.
The consequences differ depending on how a person tampered with the hardware. Anyone who physically drilled out or destroyed the LED has lost camera functionality permanently; there is no ordinary path back to a working camera. Anyone who used the far more common method of covering the LED with a sticker or similar material will regain camera function as soon as the covering is removed, since the underlying hardware is intact.
Meta's account of the scale of the problem
Alex Himel, Meta's vice president of wearables, addressed the crackdown in a Thursday post on Threads, describing those who tampered with the hardware as a "tiny minority of people." He said Meta intends to "keep rolling out updates like this to make sure the capture LED can reliably alert bystanders when photos or videos are being captured."
That framing sits alongside a broader public relations effort. Meta's chief technology officer, Andrew Bosworth, used an Instagram ask-me-anything story on Monday to argue that the glasses represent, in his words, "the most privacy-forward camera" available to consumers. Bosworth pointed to the recording light as a differentiator against smartphone cameras, saying, "You haven't had a camera that put a light on when it took a photo or video of somebody since you had like a VHS tape recorder in the 90s. Your phone doesn't have that, and it's the most commonly held camera on earth." Meta has also put up billboards in Los Angeles carrying the message that the glasses are "designed for everyone," and "not just the people wearing them."
The reputational problem the update is meant to address
The push comes after Meta's AI glasses picked up an unflattering nickname in parts of the internet and tech press: "pervert glasses," a label that stuck as reports accumulated of the devices being used to record people, including in intimate settings, without their consent. The recording light, intended as the primary safeguard against exactly that kind of misuse, was undermined by its own limitations: it is small, easy to miss in normal lighting or from an angle, and, as this crackdown demonstrates, was until recently possible to defeat with a piece of tape.
The stakes extend beyond individual embarrassment. In March, Meta was named in a US lawsuit filed by plaintiffs represented by Clarkson Law Firm, alleging that footage captured by the glasses had in some cases been reviewed by human contractors, including sensitive material such as nudity and financial information. The lawsuit followed reporting by Swedish outlets Svenska Dagbladet and Göteborgs-Posten that a Kenya-based subcontractor's staff had reviewed private footage recorded on customers' smart glasses as part of Meta's AI training pipeline. The United Kingdom's Information Commissioner's Office opened its own inquiry into the matter. The lawsuit contends that Meta's marketing language, including phrases describing the glasses as "designed for privacy, controlled by you," misrepresented how thoroughly private that footage actually stayed. For an East African audience, the involvement of a Nairobi-based content review operation places the region directly inside a dispute usually framed as a purely American consumer-privacy story.
A narrower fix for a broader problem
Closing the LED loophole addresses one specific technical workaround, not the underlying dynamic that makes the light an incomplete safeguard. A functioning indicator light only protects a bystander who happens to notice it, recognizes what it means, and can identify that the object on someone's face is a camera rather than an ordinary pair of glasses in the first place. None of those conditions is guaranteed in a crowded room, at a distance, or among people unfamiliar with what Meta's product line looks like. The update also does nothing to change what happens to footage after it is captured, which is the part of the process at the center of the pending litigation.
Bosworth's comparison to smartphone cameras, which carry no equivalent indicator, is factually accurate but does not resolve the core objection critics have raised. A phone is a general-purpose device that people recognize on sight and generally understand can record; glasses are marketed and worn specifically to blend into the face and disappear from a stranger's notice, which is a different design goal with different consequences for the people being recorded, not the person wearing the device. The absence of a comparable safeguard on phones is a reason to want better norms across the industry, not a justification for treating a small, defeatable light as sufficient on its own.
Meta faces an ongoing technical contest here rather than a solved problem. Each fix closes a specific workaround, and each fix creates pressure to find the next one, whether through new hardware modifications, software exploits, or simpler tricks that have not yet been reported publicly. The company has said it is also removing advertisements for third-party services that offer to modify the glasses to defeat the LED, and has indicated it is prepared to pursue legal action against people or businesses offering such modifications. Whether that combination of enforcement, litigation risk for Meta itself, and public messaging changes the underlying behavior, or simply pushes it further underground, is likely to become clearer as the company continues rolling out updates in this area.
Comments