Over the past year, in three investigative stories and a narrative podcast episode, The Washington Post revealed that police across the country are routinely violating municipal codes, discarding traditional investigative standards and putting the freedom of innocent people at risk through irresponsible use of artificially intelligent facial recognition software.
Our reporting found that police are treating the software “match” recommendations as evidence of guilt, with little or no investigative follow-up – despite zero scientific testing on the reliability of these tools and known wrongful arrests.
Based on our interviews with dozens of sources and extensive public records requests, we identified several troubling and previously undocumented patterns in the police use of facial recognition:
At least 15 police departments spanning 12 states arrested suspects identified through AI matches without any independent evidence connecting them to the crime — in most cases contradicting their own internal policies requiring officers to corroborate all leads found through AI.
Police often “confirmed” the accuracy of facial recognition matches with nothing more than their own visual comparison of faces or through an eyewitness photo lineup. Identification experts say both tactics can taint investigations because they could lead officers or witnesses into believing that someone who is only a “look alike” is in fact the culprit.
Authorities routinely failed to inform defendants about their use of facial recognition software, denying them the opportunity to contest the results of an error-prone technology and potentially violating a Supreme Court ruling that said defendants must be provided information that could be material to their defense.
Police in Austin and San Francisco repeatedly skirted local bans on the technology by covertly asking neighboring law enforcement agencies to run facial recognition searches on their behalf.
The stakes of this frightening problem were made clear in our story detailing how a botched investigation by St. Louis police led to the wrongful arrest and 16-month detention of Chris Gatlin. He is one of at least eight people wrongfully arrested in the United States after being identified through facial recognition, including two whose stories were first reported by The Post.
