You are on the train. The person next to you wears an ordinary pair of glasses. Frame, lenses, nothing unusual. Except there is a camera in the frame. And an AI that listens, watches and explains on request what it sees.
You did not agree. Nobody asked you. You may not even know.
Europe's privacy regulators have now thought about exactly this situation.
What the data protection board said
The European Data Protection Board (EDPB, the body of the EU's national privacy authorities) published a statement on smart glasses on 13 October. Important upfront: it is guidance, not a law and not a ban. It shows how the authorities intend to apply the GDPR (the EU privacy rules) to these devices.
The key points, as far as the statement allows:
- Responsibility lies with manufacturers and AI providers, not with the person wearing the glasses. Putting them on should not make you answerable for everything.
- Facial and voice recognition count as high risk in public. The board speaks of questions society still has to answer.
- Bystanders can hardly consent. How could they, if they cannot even spot the camera? That makes the data processing hard to justify legally. Especially if the footage later trains AI models.
- Visible or audible signals (such as a recording light) are recommended, so others notice something is running.
- Sensitive places such as doctors' offices, schools or rooms with children call for extra care.
What the statement does not say
A second look pays off here. Some media coverage sharpens the story: it speaks of a "ruling", or of a mandatory light that shuts the camera off when you cover it. The board's official summary does not back that up. It is preliminary guidance, not a rule.
And no, it does not mean every AI glasses model is illegal. Glasses that give you directions or translate a conversation are something else than glasses that recognise faces in a crowd. The difference lies in what happens to the recordings.
Why this is a real problem
You hold up a phone visibly. A pair of glasses, not so much. That is the whole point. Photography was always possible, but you could see it. Here the recording becomes a side effect, and the AI analyses it on the spot.
To be fair, the devices have real benefits. For people with impaired vision, glasses that describe their surroundings can change a lot. A reflex ban helps nobody. The question is who follows the rules, not whether the technology may exist.
What you can do
As a wearer:
- Ask before you film. When in doubt, switch it off. Politeness beats any legal situation.
- Read the settings. Whether recordings go to the cloud and whether they train AI often appears in the app's privacy options. You can frequently switch it off there.
- No use in changing rooms, medical practices, schools. Full stop.
As a bystander:
- You may ask whether the glasses are recording. That is not an insult.
- If you suspect misuse, you can contact your national data protection authority. It costs nothing.
Bottom line
The statement does not solve anything in one go. But it shows the direction: whoever builds devices that can capture others unnoticed should not count on the wearers sorting it out.
Treat the thing like a camera you cannot see: better to ask once too often.
Before you buy your next pair of AI glasses, check the privacy settings. Ten minutes well spent.