Meta is cracking down on covert recording with its Ray-Ban Meta AI glasses, even as it reportedly tests a prototype which may raise even more privacy concerns.
In a blog post this week, Meta said it is updating the second-generation smart glasses so the camera will shut off if the device detects the LED that lights up during recording has been tampered with or destroyed. The glasses already disable the camera when the LED is covered.
Meta said in the blog post that a blinking LED is an appropriate visual warning to deter covert photography, arguing that a camera-shutter sound that’s loud enough for people nearby to hear would not be practical for its glasses.
Still, at the heart of the privacy debate is how easily smart glasses, like those made by Meta, allow someone to record another person. The wearer must manually activate the camera with a button on the glasses arm or say aloud “Hey Meta, take a photo or video.”
The LED recording light, which has been a standard feature on Meta’s line of smart glasses since they were first released in 2021, was meant to address that concern. But critics have questioned how effective it has been, partly because some people don’t recognize what the blinking light means or can’t see it well in the daytime—also because some users have found ways to disable it.
Meta for its part said in its blog post that week that it is removing Facebook Marketplace listings for people offering to disable the LED on the glasses and may ban accounts or pursue legal action against people providing these services.
“More and more people use our AI glasses because they’re genuinely helpful in everyday moments like listening to music, getting live translation while traveling, or making a call hands-free.The people who use them and those around them need to trust them. That’s why we built privacy into our AI glasses from the ground up.” said Meta spokesperson Dina El-Kassaby. “We will keep strengthening our protections as our glasses become even more capable.”
Pressure has been building for Meta to tighten its safeguards. Earlier this year, the company was named in a lawsuit alleging that intimate moments captured by users’ smart glasses were later viewed by workers in Kenya who were reviewing the material to help train Meta’s AI models.
“People changing clothes, using the bathroom, engaging in sexual activity, handling financial information, and conducting other private activities inside their homes that no reasonable consumer would ever expect a stranger to watch,” the complaint states, according to the Wall Street Journal.
Photos and videos captured by a user remain private unless they choose to share them. However, when media is shared with Meta AI, contractors may sometimes review it to improve the product. This material is filtered to protect privacy and remove identifying information.
At the same time, the Financial Times reported this week that the company is testing prototype “super-sensing” glasses that would collect continuous audio and take photos every few seconds, allowing users to later query the glasses’ AI about what they saw or heard.
Executives have discussed not activating the LED while those features are in use, the FT reported, although the plans could still change. Raw footage and audio would also not be stored by Meta or be able to be accessed by the user.
“Our approach has been to develop new technologies that will help people throughout their day with privacy built in from the ground up. This work includes projects like our Aria research glasses that we showed at Connect, which uses privacy protective technologies to help people without capturing photos and videos the way traditional cameras work. While we don’t comment on internal prototypes, we’re committed to getting our glasses right because they need to be loved by both people wearing them and those around them,” said El-Kassaby, the Meta spokesperson.
Mark McCreary, a partner and chief artificial intelligence and information security officer at law firm Fox Rothschild, told Fortune that the anti-tampering safeguard on the smart glasses was a positive move. But he said it appears at odds with the Financial Times report about Meta’s super-sensing prototype.
“I mean, a cynic could say, ‘Don’t look at the fire. Look over here.’ This could be a reaction to what’s happening with the reporting about their potential new product,” McCreary said.
He added that the company’s advertising-driven business model heightens those concerns.
“I think we’ve all seen the different times over the years where Meta has been a little fast and loose with the use of their customers’ personal information,” he said. “They’ve built an entire business where 90 plus percent of their revenue comes from advertising, knowing everything they can about you and me, and then selling that to companies that will advertise to you and me.”
AI glasses add another privacy complication because the wearer may have agreed that material they choose to share with Meta AI can be reviewed or used to improve its products, while bystanders appearing in their footage or photos may not have given consent.
How consumers respond may depend on whether they view the use of photos and audio as more invasive than how other tech companies already collect massive amounts of their data.
“It’s unknown,” McCreary said. “We’re going to have to wait and see how much people feel there’s an ick factor there—or how much we’re past it and privacy is dead.”
