Sector · Law
In court the glasses are usually the evidence, not the tool
Every other sector here shows glasses helping somebody work. This one shows courts confiscating them at the door, and a witness whose entire testimony was thrown out because a voice was answering for him.
A judge in London noticed that a claimant paused for a long time before every answer. Defence counsel and the Lithuanian interpreter said they could hear a voice coming from the glasses he was wearing. He said it was ChatGPT on his phone; the judge did not believe him, found he had been coached in real time, and discarded his testimony as unreliable and untruthful. That single hearing is why this sector exists, and it is also why the next five cases are bans.
The bans, in order
A US federal court in Wisconsin added smart glasses to its prohibited-items list in February 2025, almost a year before the wave. Philadelphia's courts barred any recording-capable smart glasses, prescription or not, from every courthouse from 30 March 2026, with contempt and arrest on the table, reasoning that the risk is intimidation of witnesses and jurors. Montgomery County followed in May. In England and Wales, HMCTS said Meta glasses will be taken at the door and returned on the way out. Phones are still allowed, because a phone recording is visible; glasses are not, because they are not.
Brazil arrived by a different road. The joint CNJ and CNMP resolution of 24 September 2025 regulates audio and video capture in hearings and jury trials without naming glasses at all, which is exactly why it covers them. And for the 2026 elections the TSE spelled out that wearables count as recording devices under the 1997 law: you may carry them to the polling station, but they stay outside the booth.
What is left for a builder
Two research cases point the other way. At Louisiana State University a team built a forensic method for the Ray-Ban Meta ecosystem and showed that some identifiers and traces survive deletion, unpairing and a factory reset, publishing a parser to automate the extraction. If you sell or deploy these devices, that is the sentence to read twice.
The other is a warning about image quality. In Hannover, two doctors photographed autopsy findings with Google Glass and with a DSLR on the same four bodies. Six experts rated 159 images: the Glass photos scored significantly worse on sharpness, colour, brightness and framing, and each Glass examination took an average of 42.6 seconds longer. Hands-free capture is not the same as usable capture, and forensic work is where the difference gets tested.