The Ham Sandwich That Wouldn't Indict

There’s a saying in American legal circles that a grand jury will indict a ham sandwich. The bar is that low. So when a grand jury in Ohio looked at footage of a bloke smashing up a Flock surveillance camera, and declined to charge him, that’s not a small thing. That’s someone throwing the sandwich back.

I’d never heard of Flock cameras until this story crossed my feed, but the shape of it is familiar. Automated licence plate readers, sold to local councils and police departments as a crime-fighting tool, quietly multiplying across towns with no public vote, no real debate, just a rollout. Sound familiar? We’ve got our own version of this conversation here, slower and less dramatic, but the same instinct is at play: more cameras, more number plate recognition, more “safety” infrastructure that nobody asked for and everybody just sort of wakes up to one day. I drive past a mobile speed camera trailer near Officer most weeks and think about how normalised this stuff has become without much of a conversation about where the data goes or who gets to look at it.

The thing that got me about this particular story wasn’t the smashed camera. It was how they identified the guy: stitched together from nearby surveillance footage, a credit card transaction, and a supermarket rewards account. Not the Flock camera itself, mind you, something else entirely, a loyalty card he swiped for a sandwich. That’s the bit that should worry people more than the camera on the pole. We’ve built a world where buying a sub can place you at a crime scene. I’ve got a coffee loyalty app on my phone that I signed up for because it got me a free flat white every tenth visit. I never once thought about it as a location tracker. It is one now, apparently, whether I like it or not.

I don’t think the “nothing to hide” argument has ever held up, but it’s worth restating why, because people still trot it out at barbecues like it’s self-evident. The problem was never about hiding wrongdoing. It’s that the definition of wrongdoing moves. Protesting a war, complaining about a fine, being at the wrong intersection at the wrong time when a licence plate reader has a bad day and flags your car as stolen. None of that requires you to have done anything wrong. It just requires a system that’s confident and occasionally, sometimes badly, mistaken. There have been enough stories now of police pulling guns on innocent people because a camera misread a plate that “just trust the algorithm” should have been retired as an argument years ago.

What I find genuinely interesting, and I say this as someone who finds AI fascinating even while I’m wary of it, is that the tools getting deployed here aren’t even that sophisticated. They’re pattern-matching systems bolted onto local government budgets, sold with the same confident marketing language you’d expect from a startup pitch deck. Public safety. Peace of mind. Nobody in that pitch deck mentions false positives, or a bored cop running plates on an ex-partner, because that’s not a bug they want printed on the box.

I don’t have a tidy conclusion here, and I’m suspicious of anyone who does. Surveillance genuinely does solve some crimes sometimes. It also genuinely gets used to stalk, harass, and occasionally to ruin an innocent person’s week over nothing. Both of those things are true at once, and pretending otherwise is how you end up either paranoid or complacent, neither of which is a great place to legislate from. What I do know is that a jury of ordinary people, given the choice, decided that dismantling one of these things with a hammer wasn’t worth putting someone through a trial for. That’s not nothing. That’s a room full of strangers looking at the evidence and quietly saying: we’re not sure we want this either.