Noticias

Flock Safety builds the automated licence plate readers that sit on poles at the entrance to a neighbourhood, photographing every car that passes and logging the plate with a time and a place. According to documents obtained by 404 Media, the company pitched something considerably more ambitious: putting that same capability inside roughly 350,000 rideshare and delivery vehicles, through a partnership with the dashcam maker Nexar.

The most important fact comes first, because it is the one most likely to be lost in the retelling. Flock told 404 Media the partnership was never executed. Nothing described here is running in the cars you ride in. Uber, Lyft and Nexar did not respond to the outlet's requests for comment, so what the drivers would have been told — or whether they would have been told anything — is not documented anywhere public.

So why does a plan that never happened matter? Because of where it was presented. The pitch appears in a deck shown to the Georgia Office of the Attorney General. A company does not walk an idea into a state law enforcement office to think out loud; it does that when the idea has become a product with a price. The proposal had left the whiteboard.

What it would have changed is the shape of the network rather than its nature. A fixed camera has a fundamental limitation that anyone trying to avoid it eventually learns: it is in one place. Its coverage is a map of intersections, and the gaps between them are real. A camera bolted to a car that drives for a living has no such gaps. It goes where the fares go, which is everywhere, all day, including streets no municipality would ever authorise a pole on.

The other shift is who pays. Fixed readers are bought by police departments and city councils, which means there is a budget line, a public meeting, and sometimes a vote. Cameras riding inside private vehicles are paid for by the rideshare economy. The surveillance expands without anyone having to approve an expansion.

It is worth being precise about what these systems record, because the popular description — 'they read licence plates' — undersells it. Each read is a plate, a timestamp and a location. Individually that is unremarkable. Accumulated over months and joined together, it is a movement history: the address you leave at eight, the clinic you visit on Thursdays, the house you park outside overnight, the demonstration you attended. None of that requires anyone to have suspected you of anything.

Bystanders are captured on exactly the same terms. A pedestrian has no account to close, no setting to change and no relationship with the company holding the footage. The consent model that covers a driver installing a dashcam in their own car does not extend to the street it points at.

There is a version of this story that overstates it, and it is worth naming so you can recognise it. Flock says the deal did not go ahead, and there is no evidence in the reporting that it did. If you read that your Uber is currently scanning plates, that claim is going beyond what is documented.

The version that holds up is narrower and, in a way, more useful. A surveillance vendor looked at the limits of a fixed network, saw private commercial vehicles as the way past them, built a proposal around a dashcam partner, and took it to a state attorney general. Whether this particular attempt shipped tells you less than the fact that it was drawn up at all.

What can you actually do about it? Honestly, very little as an individual, and anyone selling you a personal fix for this is selling something. Licence plate reading happens outside your device, outside your accounts, and outside your consent. The decisions that matter are made where these systems are bought: council meetings, procurement records, state contracts, and the public records requests that make them visible. That is unglamorous, and it is the only lever that has ever moved on this.