Mashable reported that a US agency has barred its staff from using Meta smart glasses. Whatever the reasoning there, it raises a question most wearers have never asked and most people around them cannot answer: what do smart glasses actually capture, and can you tell?
The hardware
Current camera glasses contain a forward-facing camera, several microphones, a small battery, and enough processing to handle a wake word. Some add a display; most do not.
They are not recording continuously. Battery capacity in a spectacle arm is tiny, and continuous video would exhaust it in well under an hour. Recording is triggered by a button or a voice command, and clips are typically short.
The microphones, however, are always listening in the same sense a smart speaker is: locally, in a rolling buffer, waiting for a wake phrase. That audio is not transmitted until the wake word fires.
The indicator light is the whole social contract
Every mainstream pair has a small LED that lights when recording. It exists because regulators in several jurisdictions effectively required a visible signal, and it is the only thing telling a stranger they are on camera.
Its weaknesses are obvious once stated. It is small, it is easy to miss in bright sunlight, it is invisible from behind or in profile, and it means nothing to someone who does not know what it signifies.
Manufacturers have added tamper detection so that covering the light disables recording, which is a genuine improvement. It still relies on people knowing to look.
Where the recordings go
Clips are stored on the glasses, then synced to a phone app, and from there frequently to the manufacturer’s cloud service depending on your settings.
The setting worth finding is whether captured media, and any voice interactions, are used to improve the company’s products. On several platforms that is on by default and is a separate toggle from cloud backup.
If the glasses offer AI features that answer questions about what you are looking at, then images are being sent to a server to be analysed, and that is a different data flow again from recording a clip.
The legal position is not what people assume
Photography in a public place is broadly lawful in the UK, the US and much of Europe, and smart glasses do not change that.
Audio is the exception, and it is the one that catches people. Several US states require all parties to consent to a recorded conversation. Recording audio changes the analysis substantially in a way that photographing a street scene does not.
Private property is a separate matter. Shops, gyms, hospitals, schools, courts and workplaces can and do prohibit recording regardless of the general law, and changing rooms and toilets are protected everywhere.
Under the GDPR, a purely personal recording is generally exempt. Publishing footage of identifiable people, or using it for anything organisational, is not, and that distinction is where most wearers would come unstuck.
If you wear them
- Say so. One sentence when you sit down with people removes the entire problem.
- Take them off indoors in places where a phone camera would be unwelcome. The test is not whether you are permitted, it is whether people would object if they knew.
- Turn off product improvement sharing unless you actively want it.
- Check the workplace policy before wearing them to work, particularly around anything confidential.
If you are around them
Look for the light, and ask if you are unsure. Asking is reasonable and most wearers answer straightforwardly.
The honest summary is that these devices are less capable than the anxiety suggests and less obvious than the design implies. They are not silently filming your day. They can capture a moment you did not agree to, and the only reliable safeguard is a norm of saying so, which is still being negotiated.
Join the discussion