Metas Smart Glasses are often hard to distinguish from ordinary glasses on the outside, but they do have an integrated mini camera. As photos or videos are taken, an LED on the front flashes to alert people around. In a blog post, the company announces new measures against the manipulation of this recording LED.
Read more after the ad
Meta writes that since the second generation of smart glasses, the camera is automatically deactivated as soon as the system detects that users are covering the LED. In a current update, Meta is introducing a new protection mechanism: In the future, the camera will also be deactivated if the LED has been physically manipulated or destroyed.
With this, Meta is responding to reports about providers (YouTube video) who offer to permanently deactivate the LED online for a fee. The company announces that it will remove corresponding advertisements and offers from its own marketplaces, block accounts involved and also take legal action against such providers of manipulation services outside of its own platforms.
Smart Glasses: Worries about the next expansion stage
Meta repeatedly refers to the recording LED to address privacy concerns. In fact, bystanders often do not notice that they are being filmed. Be it because of strong sunlight or because they simply don’t expect a camera to be built into ordinary-looking glasses. From the perspective of data protection critics, Meta and other camera glasses manufacturers are working to normalize ubiquitous recording technology.
Once this is widely accepted, it is feared that additional functions could be gradually introduced that further undermine data protection. There have long been reports that Meta is considering introducing facial recognition for its smart glasses. A corresponding code was discovered in the glasses’ companion app, but it was not activated. Meta is also said to have tested facial recognition from a Pentagon supplier.
The Financial Times reports today that Meta is testing prototype glasses that can capture the wearer’s everyday life almost completely. With the feature called “Super Sensing,” users could use AI to query what they saw or heard or reconstruct their day. According to the report, Meta executives are considering not enabling the recording LED while using this feature and are discussing other privacy mechanisms instead. According to the informants, the function could also be activated on existing Meta glasses via a software update. There were reports of a feature called Super Sensing over a year ago.
Read more after the ad
When private recordings end up with AI service providers
In its own blog post, Meta also addresses the question of what happens to the photos and videos that users take with the smart glasses, apparently responding to a data protection scandal from the spring.
At the time, Svenska Dagbladet reported that employees of a Kenyan meta-service provider had access to highly sensitive recordings of smart glasses users, including videos of sexual acts, as part of an AI data annotation. Meta then ended its collaboration with the service provider and more than 1,000 employees lost their jobs.
There has been no transparent processing of the case since then, which is unlikely to strengthen trust. When asked by heise online, a Meta spokesman said at the time that Meta sometimes uses subcontractors for content that users share with Meta AI, who evaluate this content in order to improve the functionality of the smart glasses. This is clearly stated in the guidelines. Apparently this processing cannot be switched off completely because Meta AI is deeply integrated into the smart glasses.
The blog post reassures Meta users by saying that only they can see their photos and videos. Unless you share them with friends, publish them on social media – or use them with Meta AI.
(tobe)
