The conversation around smart glasses privacy has been stuck in a familiar groove for years: does the recording light work, can people tell when they’re being filmed, and what happens to the footage. But a pair of intertwined developments out of Meta this week suggest the conversation needs to shift entirely. Because the next generation of smart glasses isn’t just about whether the camera indicator lights up — it’s about whether the camera ever really turns off.

Earlier this week, Meta started pushing a mandatory update to existing Ray-Ban smart glasses that disables the camera if it detects the LED recording light has been tampered with. It was a direct response to a cottage industry of modders offering paid services to physically remove or disable the indicator, which had been used to record people without their knowledge. On the surface, it looked like a responsible move — a company acknowledging a genuine safety flaw and engineering a fix.

Then came the follow-up. The Financial Times reported that Meta is actively developing “super sensing” smart glasses with microphones and cameras that stay active continuously, snapping photos every few seconds and constantly listening for AI queries. And according to people familiar with the current plans, the LED indicator that just got so much attention? It won’t be on during those always-on sessions. Not because of tampering. By design.

The reasoning Meta has floated internally is that if the LED blinked every time the AI features were running — which could be all day — people would stop noticing it anyway. The light would become background noise, defeating its purpose as a signal of active capture. In a 2025 whitepaper, the company argued that reserving the LED for “active capture” moments (when the user is consciously saving a photo or video) would preserve its meaning. The always-on sensor mode would simply operate without it.

There’s a logical thread here that’s worth following. If a device’s sensors are always running — not recording video for later playback, but extracting metadata for AI queries in real time — does a recording light even make sense as a privacy signal? The light was designed for a world where cameras turned on and off explicitly. An always-on system breaks that model completely. The wearer isn’t pressing record. The glasses are just… glasses. But they’re glasses that can answer “what did that person say ten minutes ago?”

Meta says the raw footage and audio in this “super sensing” mode won’t be stored or made available to the user. Instead, metadata gets extracted and uploaded for the AI to query. Proponents argue this is less privacy-invasive than keeping full recordings. Critics counter that the distinction matters little to the person who was unknowingly monitored all afternoon.

The timing couldn’t be more awkward. Meta just finished closing one privacy loophole — the tampered light problem — and now finds itself defending a very similar gap in an entirely new product line. The company’s statement emphasizes being “committed to getting our glasses right because they need to be loved by both people wearing them and those around them.” But the gap between that ambition and the engineering reality is getting harder to ignore.

What makes this moment different from earlier smart glasses controversies is the fundamental shift in what the hardware does. Earlier debates were about surreptitious recording — a known bad act enabled by a broken indicator. The always-on question is harder because it challenges the core assumption that sensors should have an off state at all. If the glasses are always sensing, and there’s no light to indicate that state, the social contract around wearable cameras has to be renegotiated from scratch.

Source: The Verge and 9to5Google