Detroit mother files lawsuit over facial recognition arrest

Our Take

Porcha Woodruff's story, reported here by WDIV's Demond Fernandez for NBC News, is exactly the nightmare scenario privacy advocates have been warning about for years. A facial recognition match — not a human witness, not solid detective work — put officers at her door, and she was hauled away eight months pregnant over a robbery she had nothing to do with. The algorithm was wrong. The consequences were very real.

This case gets cited constantly because it's so clean an example of automation bias: once a computer spits out a name, humans downstream treat it as confirmed fact instead of a lead that needs corroboration. Detroit PD had already been caught in prior wrongful arrests tied to the same technology, and yet the pattern repeated. That's the danger of letting any surveillance system — facial recognition today, ALPR networks tomorrow — become the de facto probable cause instead of a tool that still requires real verification.

We track this same mission-creep with license plate readers because the underlying problem is identical: a machine makes a guess, a database logs it, and nobody stops to ask whether the guess deserves the trust it's given. If Flock cameras or similar systems are logging plates in your area, check our map to see what's already watching, and visit our take-action page if you want to push back on unaccountable surveillance before it puts the wrong person in handcuffs.