MOSCOW (Bernama-Sputnik/RIA Novosti) – US law enforcement agencies are using artificial intelligence (AI) tools as a shortcut to finding and detaining suspects without other evidence, ignoring obvious clues, reported Sputnik/RIA Novosti.
The Washington Post newspaper on Monday reported that fifteen police departments in 12 states detained suspects identified by AI algorithms without any independent evidence linking them to the crime, adding that law enforcement agencies in most cases contradict their own internal policies, which require police officers to back up any evidence found with the help of AI.
According to The Washington Post investigation, at least eight people were mistakenly detained by US investigators using facial recognition technology. In some cases, the police did not check the suspects’ alibis or false witness statements, and in others, they ignored fingerprints and DNA traces that pointed to other people, the report read.
The report noted that investigators did not even take into account obvious differences in the appearance of criminals recorded by surveillance cameras and the suspects identified by AI. There was a case when law enforcement officers detained a woman who was seven months pregnant, accusing her of carjacking, despite the fact that nothing indicated that the actual criminal was pregnant.
The real scale of the problem is most likely much larger, the newspaper reported, adding that US investigators are not legally required to provide data on the use of AI facial recognition.
