There’s been a lot of social blowback against camera-enabled smart glasses this year, particularly those from Facebook-maker Meta. A post on Tuesday from the company addresses privacy concerns raised by me and others over the last year, and there’s also a new mandatory firmware update for some Meta glasses that promises to deactivate the camera if any tampering of the small recording LED light is detected.

That’s actually been a thing: There have been services that would mod and deactivate the recording light on the glasses, turning them into even more stealthy recording devices. Meta’s post promises that this update will end the problem of recording people without their knowledge.

But Meta hasn’t yet solved any of the other issues with camera smart glasses, including concerns about how or when the glasses might be recording in public. The LED camera indicator on existing glasses is more than phones offer, but the indicator light is often hard to spot in daylight, and Meta’s smart glasses are so normal-looking now that it’s hard to tell they have cameras at all.

A wave of bans on smart glasses in certain places, such as New York courtrooms and parts of cruise ships, has already begun. Public policies on glasses are likely to get more intense. Some glasses-makers are offering camera lens covers, and Meta’s Andrew Bosworth admitted during a Meta glasses launch a few weeks ago that he’s interested in making camera-free glasses down the road.

Privacy remains an issue, especially as capabilities increase

Meanwhile, Meta’s own privacy policies for how photos and AI voice interactions are handled in the Meta AI app that’s used to connect with glasses are still murky. Meta tries in its latest post to lay out clearly how personal photos aren’t shared with Meta’s cloud services by default, but it’s still easy to use the glasses to analyze photos with AI in ways that would end up cloud-sharing important information. 

The challenges are only going to expand. Right now, the batteries in Meta’s glasses can’t handle continuous recording for more than an hour or so, but the Financial Times reports the company is trying to develop future “super sensing” glasses that record more often for longer periods of time. Based on where Qualcomm’s existing wearable chipsets are heading, with better battery life and AI performance, that’s exactly where future smart glasses are heading. Many of these devices want to be wearable extensions of our memory, and that means staying on even longer. Meta’s Orion prototype AR glasses I tried a few years ago are a sign of those ambitions.

Meta is also expected to introduce facial recognition tech into its glasses, a capability I saw demoed on TCL glasses with Qualcomm chips just a few weeks ago.

And also, Meta’s own rules on privacy and AI keep shifting, often suddenly: take for example Meta’s new AI image tools that can deepfake anyone’s Instagram account selfie.

As Google and likely Apple enter this market soon, it’s a privacy question that needs to be handled now. And it won’t just be on glasses. More constantly recording pins, pendants, watches and other devices hooked into AI are emerging seemingly every day. I’ve worn many of them. Augmented reality devices, like Snap’s upcoming Specs glasses, will have these types of camera-enabled AI sensing features too, to layer experiences realistically onto the world around us.

And while Meta’s glasses don’t actually record all the time now as much as people think they do, odds are they will soon. We can’t wait for glasses to truly develop that type of always-on, AI-enabled, super-sensing function for these privacy concerns to be better addressed.