The Dutch privacy watchdog has warned consumers against buying smart glasses after tests found serious AI errors, security flaws and potential privacy risks.
The Dutch privacy watchdog has advised people not to buy smart glasses after testing found the devices could give dangerously incorrect advice, make false claims about contacting emergency services and expose users to security risks.
The Autoriteit Persoonsgegevens (AP) tested six pairs of smart glasses available in the Netherlands, ranging in price from less than €100 (£84) to more than €500.
The authority did not identify the manufacturers or say whether Meta’s smart glasses were among those tested.
During a simulated emergency, one pair of glasses repeatedly claimed that an ambulance was on its way after being asked for CPR advice, despite emergency services never being contacted.
The AP also found potentially serious errors involving child protection, with one pair providing dating advice to an adult who said they were romantically interested in a minor.
The glasses also misread instructions on medicine packaging and incorrectly interpreted severe weather warnings.
While the watchdog stopped short of calling for smart glasses to be banned, it said the findings demonstrated that consumers should be cautious.
The AP’s investigation comes amid growing concerns about the privacy implications of wearable devices equipped with cameras and artificial intelligence.
The authority warned that some glasses could be used to secretly record people, with concerns that recording indicator lights could be obscured while the devices continued functioning.
In some cases, the AP found that people nearby could connect to the glasses without permission, view recordings or even operate the camera remotely.
It also warned that data recorded by smart glasses was frequently transferred outside the European Union.
Meta has previously promoted its smart glasses as useful tools for people with visual impairments, with AI able to describe surroundings and read text.
The AP acknowledged that such technology could be valuable, but said reliability was particularly important for people who depend on it for assistance.
The findings come after reports that outsourced workers had been able to view sensitive material recorded using Meta’s smart glasses.
Meta said at the time that some content could be reviewed to improve its AI and that it took data protection seriously.
Dutch watchdog warns against smart glasses







