The Exemption Would Need Its Own Database

Who counts as disabled is not one question, it is at least four. Who is the user, a category that splits across diagnosis, functional limitation, self-identification and fluctuating conditions that no single form captures well. Which device counts, when a clip-on camera, a prescription frame and an audio-only pair can look identical from across a room. Which function is protected, since scene description, a live video call, and saving footage to post later all run through the same lens with wildly different risk. And, hardest of all, what counts as proof, the thing a retailer, a school, or a security guard is supposed to check before deciding the exemption applies.

The first three are hard and solvable with enough drafting patience. The fourth is where the design breaks.

There are only two ways to make proof checkable in the moment. Either the person carries a disclosure token, a card, a prescription, a registration number, something they show on demand, which means disclosing disability status to a stranger at a counter as the price of using a consumer product. Or a venue, retailer, or officer checks a name against a central list, which means somebody built a database that maps identity to disability status, to device ownership, and, the moment that list gets used at a specific location, to where that person was standing.

A policy meant to protect a group's privacy would need, as its enforcement mechanism, a record of exactly the population it exists to protect. That record is more sensitive than almost anything a bystander notice rule was ever trying to prevent a camera from capturing. It does not matter how good the access controls on that database are. Its existence is the harm, independent of who eventually reads it.

The Convention on the Rights of Persons with Disabilities protects disabled people's privacy on an equal basis with everyone else, not on a reduced basis in exchange for access to a product. An exemption whose verification step requires more disclosure than a non-disabled buyer ever has to make fails that standard on its own terms, regardless of how well-intentioned the exemption was.

The stakes are not abstract. The WHO and UNICEF estimate more than 2.5 billion people need at least one assistive product and nearly a billion do not have access to one. A verification system heavy enough to satisfy a retailer's compliance department will filter out exactly the people least able to clear it: someone with an undiagnosed or fluctuating condition, someone who cannot afford the specialist appointment that would produce the paperwork, someone visiting from a country whose documentation nobody at the counter recognises.

We have argued elsewhere in this series that regulation should attach to the mode of use rather than the object, and this is the sharpest reason why. A rule that constrains what durable capture and biometric identification are allowed to do, regardless of who is wearing the glasses, needs no registry, because it never asks who the wearer is in the first place. It asks what the device just did.

That also matches the honest limit of what a tool like ours can see. NearLens reports compatible smart-glasses signals observed nearby. It was built without the ability to read anyone's identity, and that omission was deliberate. Any accessibility policy that ends up needing a list of names to function has built the one thing this entire category should have learned, by now, not to build.