Grandmother falsely accused of theft after alleged AI facial recognition match sues

Our Take

ABC News lays out a nightmare scenario: a grandmother held for half a year on charges tied to an algorithm's guess. We don't know every detail of the match or the investigation yet, but the broad shape of this story is depressingly familiar — a black-box system flags someone, humans treat that flag as gospel, and a person's life gets wrecked before anyone bothers to double-check the machine's work.

This is exactly the pattern we track with automated license plate readers. Flock Safety and similar systems operate on the same core assumption that facial recognition tools do: that pattern-matching software is reliable enough to drive arrests, detentions, and prosecutions with minimal human skepticism. When that assumption fails — and it fails disproportionately often for certain groups — the burden of proving innocence falls entirely on the wrongly accused, not on the vendors or agencies who deployed the tech.

Lawsuits like this one matter because they're often the only mechanism forcing transparency about how these systems actually perform in the field, since the companies selling them rarely volunteer failure rates. If you want to see how this same automated-first, verify-later mindset is showing up in your own community through license plate surveillance, check our camera map, and if you're ready to push back on unchecked deployment near you, visit our take-action page. Credit to ABC News for surfacing this case.