Facial Recognition Software: 14 Wrongful Arrests So Far

Our Take

Fourteen wrongful arrests isn't a glitch, it's a pattern. CaraComp's rundown shows what happens when facial recognition algorithms trained in sterile lab conditions get shoved into messy, high-stakes street policing. The gap between 'the software flagged a match' and 'this specific human being committed a crime' keeps swallowing innocent people, mostly without any meaningful review before handcuffs come out.

This matters to the ALPR fight because it's the same playbook. Flock's camera network and facial recognition systems both sell police departments on the promise of infallible machine certainty, then quietly shift the burden of proof onto the person wrongly flagged. A license plate misread or a flawed face match gets treated as probable cause first and fiction-checked later, if ever. Automated systems built to scale surveillance rarely get built with equal investment in catching their own mistakes.

If your town already has Flock cameras running or is considering them, that trust-the-algorithm mentality is coming for plate reads too, not just faces. Check our camera map to see what's watching near you, and use our take-action resources to push your local officials for real oversight before the next wrongful stop happens in your community.