A Computer Said He Looked Like the Thief. His Daughters Watched Him Get Arrested.
The short version: The Robert Williams case shows how a single flawed algorithmic match can upend an innocent life — a warning for every automated surveillance system, including ALPRs.
Our Take
Credit to Falsely Accused Cases for laying out the Robert Williams story with the detail it deserves. This isn't ancient history — it's 2020, in Detroit, and it's a preview of what happens when police departments treat an algorithm's guess as probable cause. A blurry still, a facial recognition 'match,' a loss-prevention contractor rubber-stamping a photo array — that's the entire chain of custody that put a man in handcuffs on his own lawn while his kids watched.
We cover license plate readers, not face recognition, but the through-line is identical: a machine spits out a match, a human treats it as ground truth, and due process gets skipped. Flock's ALPR network is already flagging cars as 'suspicious' based on partial plate reads and vague descriptors, feeding that data to hundreds of agencies with little oversight and even less accountability when the system gets it wrong. Williams got a wrongful arrest and a lawsuit that forced policy changes. Most people flagged by license plate readers never even find out they were misidentified in the first place.
That's the real danger of automated policing — not just the false positives we hear about, but the thousands we don't. If you want to see how much of this infrastructure has already landed in your town, check our camera map, and if you're ready to push back on it locally, our take-action page has a place to start.
This is DeFlock The USA’s original commentary. The video above is the work of Falsely Accused Cases, published on YouTube — full credit to the creator.