Florida Man Sues Police After Wrongful Arrest Linked to AI Error

A 52-year-old Florida man is taking legal action against several law enforcement agencies following his arrest for a crime he did not commit. Robert Dillon was accused of trying to lure a young girl at a McDonald’s restaurant in August 2024. However, the case against him was based on a mistake made by an artificial intelligence (AI) facial recognition system. According to The Guardian, the system gave a 93 percent probability that Mr. Dillon was the man seen in security footage. The charges were eventually dropped, but the experience has left a lasting impact on his life.

Mr. Dillon lives in Fort Myers, which is more than 300 miles away from the scene of the alleged crime in Jacksonville Beach. He told detectives that he had never visited that town in his life. Despite this, he was arrested at his home in front of his wife. The lawsuit, filed by the American Civil Liberties Union (ACLU) on his behalf, states that the arrest caused him serious reputational damage. Mr. Dillon now feels uncomfortable being around children, and he says that no police agency has apologized for the error.

The lawsuit claims that the police relied too heavily on the AI software rather than performing a proper investigation. The system used is called the Face Analysis Comparison and Examination System (FACES), which is managed by the Pinellas County Sheriff’s Office. According to Ars Technica, this database is used by various agencies to identify suspects. The legal document argues that instead of checking the machine’s result against other facts, the officers worked to build a case that supported the computer’s mistake.

Evidence suggests that the police had access to information that could have cleared Mr. Dillon early on. The lawsuit alleges that lead investigator Scott O’Connell ignored several important facts. For example, license plate readers showed that none of Mr. Dillon’s vehicles were anywhere near the restaurant at the time of the incident. Additionally, the photograph used for the facial recognition match was of very poor quality. It was not a clear digital file, but rather a photo of a computer screen taken with an officer’s cellphone. The lawsuit claims that this low-quality image made it much harder for the software to provide an accurate result.

Furthermore, the investigation relied on a witness statement that was likely incorrect. A McDonald’s employee identified Mr. Dillon from a photo line-up, claiming he was a regular customer. However, the lawsuit notes that it would have been impossible for him to be a regular visitor given that he lived hundreds of miles away. The legal team argues that the investigator knew this but did not challenge the witness’s claim. By leaving out this exculpatory evidence—information that shows someone is not guilty—the police moved forward with an arrest that should never have happened.

This case is not an isolated event. According to the lawsuit, it is at least the 15th time in the United States that a person has been charged or arrested due to a false identification by facial recognition technology. Experts are increasingly concerned about how these tools are used. A recent investigation by The Guardian found that the oversight of these systems is not good enough. It noted that the rapid development of AI technology is moving much faster than the ability of authorities to create rules to control it.

The implications of this case are significant for the future of policing. Nate Freed Wessler, a director at the ACLU, stated that police departments have a responsibility to make amends and to ensure that such errors do not happen to anyone else. As technology continues to change the way police work, the balance between using new tools and protecting the rights of innocent people remains a major challenge. For Mr. Dillon, the damage to his reputation remains, as his mugshot is still available online, serving as a reminder of a mistake that changed his life.