The Camera and the Recognition Service Are Sold Separately

Privacy International has argued for years that real-time facial recognition should never be coupled with body-worn cameras. The objection is to the join, not a camera by itself or recognition software in the abstract.

When both halves sit inside one procurement, a rule against coupling has something to bite on. Somebody signs for a system, the system has a specification, and the specification either includes recognition or it does not.

Consumer glasses arrive with the coupling unmade. The manufacturer can state truthfully that the product does not identify anyone, and the statement can be accurate on the day it ships. What the device does is capture and transmit, and what happens after transmission is a separate service with a separate provider, chosen later, by the buyer.

A rule requiring manufacturers not to ship recognition can be satisfied while the combination exists in practice. Assessed separately, the camera clears and the service clears. The arrangement people fear goes unassessed because no single party assembled it.

A device whose output can reach an identification service could be evaluated as though it carried that capability, because the combination is functionally possible. The relevant question includes what the hardware makes easy.

That standard has a real cost. Written broadly, it captures almost every camera manufactured today, which is unlikely to be workable. The boundary may need to come from form factor and whether the person being recorded can tell it is happening, rather than capability alone.

An assessment of combinations could examine friction as well as theoretical compatibility. A workflow that requires specialist equipment and several manual steps presents a different practical risk from one that routes a live image to a search service through a consumer menu. That distinction keeps the analysis tied to observable use rather than every possible connection.

Responsibility may also be distributed across the chain. The camera provider controls capture and signalling, the service provider controls matching and retention, and the wearer initiates the use. A rule aimed at the combined outcome would need to specify which obligation belongs at each stage instead of assuming one party assembled the whole system.