Meta closes a second privacy-light loophole on its smart glasses -- five months after a lawsuit said human reviewers, not just wearers, see the footage

Meta is rolling out a fix to its smart glasses that closes a specific, previously-documented workaround: a wearer could start recording, then cover the capture LED, and the camera would keep running -- because the software only checked whether the light was blocked *before* a recording began, not during one. "A fix for this is starting to roll out," AR VP Alex Himel said. "The camera will now stop working if the light is covered during a recording.'' It's the second time in under two months Meta has patched the same indicator light against a real-world bypass.

The first fix, in July 2026, targeted people physically modifying the hardware to disable the LED outright, and made the camera refuse to start recording if the light was already covered. That left the mid-recording gap this update now closes. Himel says the pattern continues: "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." 9to5Google reports that only a small minority of users actually try to defeat the light -- but a genuine market for accessories built specifically to beat it without detection, per reporting cited by MakeUseOf, means the minority isn't zero.

The lawsuit, filed in federal court in San Francisco by plaintiffs Gina Bartone and Mateo Canu and represented by the Clarkson Law Firm, doesn't turn on the light at all. It turns on what Meta told buyers versus what actually happens to their footage. The complaint quotes Meta's own marketing -- glasses "designed for privacy, controlled by you," with copy telling buyers "you're in control of your data and content" -- against reporting that a Kenya-based subcontractor's workers have reviewed footage including nudity, sexual activity and bathroom scenes, material the plaintiffs say they had no reason to expect a stranger would ever see.

Meta disputes the *marketing* was misleading, not the underlying practice. "When people share content with Meta AI, we sometimes use contractors to review this data for the purpose of improving people's experience, as many other companies do," the company told TechCrunch -- a defense that concedes the review pipeline exists while framing it as industry-standard. Meta has also said it blurs faces in reviewed footage, though the sourcing behind that claim disputes how consistently that actually happens. Footage shared with Meta AI feeds that pipeline with no opt-out available to wearers.

The two threads -- the indicator light and the review pipeline -- are easy to conflate and address genuinely different risks. The LED problem is about *bystander* consent: whether someone near the glasses can tell they're being recorded. The lawsuit is about *wearer* consent: whether the person who bought the glasses understood a stranger, not just an algorithm, might eventually watch what they captured. Today's fix makes real progress on the first. It says nothing about the second, which is still sitting in front of a federal judge.