---
title: Do Not Buy AI Smart Glasses, Dutch Watchdog Warns
description: The Dutch privacy regulator tested six AI camera glasses and found false emergency claims, hidden recording and security flaws. What it means for buyers.
author: Dr Marina Nani (Editor-in-Chief)
date: 2026-11-22T10:27:24.088Z
updated: 2026-11-22T10:28:22.059Z
canonical: https://www.sovereignmagazine.com/article/dutch-watchdog-ai-smart-glasses-warning
image: https://cdn.nanimediahouse.com/pexels-4469777.webp
categories: Tech News
content_type: News
region: Netherlands
publication: Rich Man Magazine
---

Someone has collapsed in front of you. Your hands are on their chest, your own heart is pounding, and the glasses on your face tell you that an ambulance is on its way. You keep going because you believe help is coming. Nobody has called for help at all.

That is what happened when the Autoriteit Persoonsgegevens (the Dutch Data Protection Authority) staged a medical emergency to see how AI (artificial intelligence) camera glasses would cope. One pair said several times that an ambulance was coming, although the emergency services had never been contacted. On 19th November the authority published its findings on six pairs sold in the Netherlands and gave shoppers its advice in unusually plain terms: do not buy these glasses for general use, at least for now.

## An ambulance that was never called

The glasses in the study look like ordinary frames. A camera and a microphone sit inside them, and you can ask a built-in chatbot about whatever is in view. The authority did not name the six products. It said only that they cost from under €100 to more than €500 and came from manufacturers on different continents, and it added that the price tag tells you nothing about how safe a pair is.

The false ambulance was the most alarming answer, and there were others. One pair gave dating advice to an adult who said they were romantically interested in a minor. The glasses made mistakes when reading the instructions inside medicine packets, misread severe weather warnings and got the safety warnings on children's toys wrong.

Monique Verdier, the authority's vice chair, said that glasses which film people without their noticing were worrying enough, and that a pair giving wrong advice when a life is at stake 'goes one step too far'. The report makes a wider point about confidence. An assistant that speaks calmly in your ear sounds as though it knows, and the tests showed that it can sound convincing while being wrong.

## A recording light that a strip of tape can hide

If you are not the one wearing the glasses, the fear is being filmed without knowing it. Manufacturers answer that fear with a small light that comes on while the camera records. According to the authority's report, only one of the six pairs had a light noticeable enough to alert the people nearby. On some pairs the light could be covered with tape and the camera carried on recording, even though manufacturers had claimed it would stop.

The software had problems too. Several pairs had security flaws that would let a stranger standing nearby connect to your glasses without your permission, look through your recordings or operate the camera. Much of what the glasses recorded was sent to servers outside the EU (European Union).

Meta, the market leader, told the Reuters news agency that it was considering a response. The company has said before that every pair with a camera carries a bright light that blinks during recording and cannot be switched off. It is not known whether Meta's glasses were among the six tested. China's TCL/RayNeo and Rokid and Taiwan's HTC, which also sell camera glasses in the Netherlands, did not immediately reply to Reuters.

## Why the regulator stopped at advice

The authority can fine companies that process personal data unlawfully, but it cannot take a product off the shelves. That decision belongs to product safety authorities. Its reach may grow, because it is preparing to take on broader enforcement powers under the [European Union's AI Act](https://www.richmanmagazine.com/article/claude-enterprise-smart-reports-employer-risk), which could give it more say over consumer products that run on artificial intelligence.

Some campaigners think advice falls short. Evelyn Austin, executive director of the Dutch digital rights group Bits of Freedom, told the Dutch news agency ANP (Algemeen Nederlands Persbureau) that the regulator had its job the wrong way round. She compared the warning to a food safety authority asking people not to eat a dangerous product instead of removing it from sale, and she said the 'peeping glasses' should be pulled from the market.

Dutch shops and courts are already moving. Hans Anders, one of the largest eyewear chains in the Netherlands, suspended sales of [Meta's Ray-Ban smart glasses](https://www.richmanmagazine.com/article/hey-meta-ray-ban-meta-smart-glasses-with-multimodal-ai-tech-in-everyday-accessories) there and in Belgium in October. Consumer and rights groups sued Meta the same month, alleging that its glasses violate privacy laws, and a preliminary hearing is scheduled for 26th November. In September the authority had warned that if you share footage of recognisable people without their permission, you break the GDPR (General Data Protection Regulation), and it urged hospitals, courts and places of worship to ban the glasses on their premises.

## The people who need the glasses to work

None of this means the technology is worthless. For people who are blind or partially sighted, glasses that describe a room or read a label aloud can change an ordinary day. Paul Arnold, chief executive of Britain's Information Commission, has a severe visual impairment. He wrote in September that the glasses help him to understand his surroundings and find his way, and in the same piece he said he had been 'shocked and appalled' by some of the ways people use them.

The Dutch authority accepts that value and turns it into an argument. The people who lean on the glasses most are the people with the most to lose when the artificial intelligence [gets a medicine label or a weather warning wrong](https://www.richmanmagazine.com/article/ecri-ai-errors-patient-care-reporting-network), so for them reliability is the whole point of the product.

Meta recently released a version of its glasses without a camera, and some commentators read that as a response to the [privacy backlash](https://www.richmanmagazine.com/article/apple-smart-glasses-siri-apps-privacy-2027).

## What you can do now

The study describes two groups carrying two different risks. If you wear the glasses, you live with the consequences of unreliable answers. If you are standing nearby, you carry the privacy and security risk without ever having chosen the product.

If you already own a pair, the tests offer three practical lessons. Keep recordings personal and do not post footage of people who can be recognised. Treat any answer about health, medicine or safety as unchecked until a person or a printed label confirms it. In an emergency, make the call yourself and hear a human voice on the line.

The authority has not asked for a ban, and the glasses will keep improving. The open question is how they become normal. They may earn their place because the lights get brighter and the answers become dependable. They may also slip into daily life simply because the rest of us stop noticing the cameras, and that second route would leave every problem in this report unsolved.
