Meta's next specs—now with extra surveillance

TL;DR: Meta is reportedly working on a new AI smart glasses prototype that’s always watching and listening, per the Financial Times. It’s not a great look for a company that’s been in a lot of hot water for privacy issues, but its existing AI glasses continue to be a runaway hit.

What happened: The pitch, based on the FT’s reporting, is a device that makes Meta’s AI more useful: By constantly capturing your surroundings, the glasses can answer immediately if you ask, “Hey Meta, what’s that restaurant I just passed?” It could even be used to “recall” your entire day.

You can probably see the privacy objections coming from orbit. While the company’s existing lineup has a small LED light that turns on when actively recording or taking photos, the light on these new frames might not activate during background capture, insiders told the FT. These controversial “always-on” features, which could be added to Meta’s existing glasses in a software update, have reportedly already sparked an internal debate around privacy and handling user data—including whether the recordings would be used to train Meta’s AI models.

Just trust us (again): Over the years, Meta has tried to assuage people that it takes privacy concerns seriously, particularly after public outcry. Just yesterday, it announced that it’ll disable the camera on its smart glasses if the system detects that the LED recording indicator has been covered or modified.

But the company’s broader privacy track record is… checkered, to say the least. Earlier this year, Meta contractors in Kenya revealed that they reviewed intimate footage recorded with Meta Ray-Bans as part of their AI data labeling work, seeing everything from “living rooms to naked bodies.” Meta’s been hit with several lawsuits since this report came out, though it claims the faces were filtered and blurred and that it gets “clear user consent” for this kind of review (although how it does this is uncertain).

Thorny, and getting thornier: Smart glasses have already stoked plenty of consternation; they’re banned from certain spaces on some cruise lines and will soon be prohibited from every New York State courtroom. With the rollout of such always-on features, Meta could find itself trying to dodge an even greater onslaught of privacy-related lawsuits, including ones related to two-party consent recording laws.

Bottom line: The growing category of “always listening and watching” devices gives many people the creeps. But if any company can cash in on them, it’s probably the one with the very popular smart glasses people are still happily wearing. —WK

Also at Meta…

• The company’s first AI image generator is out—and if you don’t want people using your photos, you might want to lock down your Instagram account.
• A contractor at Meta’s Cheyenne, Wyoming, data center flushed water contaminated with a rare bacterium into the city’s sewers—though, fortunately, city officials say it hasn’t affected the drinking water.