Law enforcement facial recognition arrests without corroborating evidence
A Washington Post investigation found that law enforcement agencies across the United States have used facial recognition software to arrest suspects based solely on AI-generated matches, contradicting their own policies requiring corroboration and abandoning traditional investigative standards. At least eight people, including Christopher Gatlin who spent over two years fighting the charges, were wrongfully arrested after being identified through the technology.
What happened
St. Louis County Transit Police detective Matthew Shute uploaded a still image from a blurry surveillance video of a train platform assault to a facial recognition program. The software generated names and photos of people deemed to resemble one of the assailants, despite the poor image quality and the suspect's face being partially obscured by a winter coat and surgical mask.
Shute proceeded to build a case against Christopher Gatlin, one of the AI-generated results, despite the city's facial recognition policy warning that results are 'nonscientific' and 'should not be used as the sole basis for any decision.' Gatlin had no apparent ties to the crime scene nor a history of violent offenses.
A Washington Post investigation found that 15 police departments spanning 12 states arrested suspects identified through AI matches without any independent evidence connecting them to the crime, contradicting their own internal policies requiring corroboration of AI leads. At least eight people were wrongfully arrested after being identified through facial recognition, including Gatlin and Miami resident Jason Vernau.
Reported impact
Christopher Gatlin, a 29-year-old father of four, was arrested and jailed for a crime he says he did not commit and took more than two years to clear his name.
Some law enforcement officers using the technology abandoned traditional policing standards and treated software suggestions as facts, with one police report referring to an uncorroborated AI result as a '100% match' and another saying police used the software to 'immediately and unquestionably' identify a suspected thief.
Index analysis
The Post reviewed documents from 23 police departments where detailed records about facial recognition use are available. Most police departments are not required to report that they use facial recognition, and few keep records of their use of the technology.
Governance controls
Pagedale's facial recognition policy warns officers that the results of the technology are 'nonscientific' and 'should not be used as the sole basis for any decision,' but officers proceeded to arrest suspects based solely on AI matches.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.