Skip to content
Privacy & Security

AI Facial Recognition Software Leads to False Arrests, Ruined Lives in Florida

An AI-powered dystopian future seems increasingly inevitable to many these days, but for some, it’s already here.
By

Reading time 4 minutes

Comments (4)

In recent weeks, two Florida cases involving wrongful arrests tied to AI facial recognition have drawn renewed attention, one through newly dropped charges and another through an ACLU-backed lawsuit. One man lost his job, car, and home as a result of spending nearly three months behind bars for a crime he did not commit, while another was falsely accused of luring a child away from a McDonald’s. Incredibly, both men said they were hundreds of miles away from the scenes of their supposed crimes when they occurred, with one of them sharing timesheets from work as evidence.

The most recent case involves Jalil Richardson, a Charlotte, North Carolina-based father of 10 who says Jacksonville police made him a suspect in a stolen vehicle case after an automated facial recognition search flagged him from surveillance footage. According to a local report from Action News Jax, the investigation began on April 2, 2025, when a victim told the Jacksonville Sheriff’s Office (JSO) that he had unknowingly bought a stolen car from a man he met at a Publix grocery store. Investigators compared parking lot footage to Richardson’s photo using Automated Facial Recognition, or AFR. Richardson’s wife, Jasmine Jackson, said officers later told the family the system returned an 85% match.

Richardson told WSOC-TV in Charlotte that he had never been to Florida in his entire life. His timecards later showed he was at work in North Carolina, roughly 400 miles from Jacksonville, when the car sale took place. By then, the case had already consumed months of his life. Richardson told Action News Jax he learned about the Florida warrant after calling police to his Charlotte home for an unrelated disturbance. He spent 33 days in Mecklenburg County custody before being extradited, then another 50 days in Jacksonville.

Prosecutors eventually dropped charges, including dealing in stolen property, grand theft, possession of a vehicle with a removed VIN, possession of a fraudulent title, and fraudulent use of fictitious identification information. However, the dismissal did not undo the damage. According to Jackson, the family was homeless after the case ended, and Richardson had lost his job, car, and stability for their children.

“It’s very traumatizing and unbelievable,” Richardson told WSOC-TV. “I lost everything . . . There was no proper investigation done to even reach out to me or to see if I was even in Florida. He just automatically put a warrant out for my arrest.”

The other Florida case now at the center of an ACLU-backed lawsuit involves Robert Dillon, a 52-year-old Fort Myers man arrested in August 2024 after being accused of trying to lure a 12-year-old child away from a McDonald’s in Jacksonville Beach. CBS News reported that the incident began in November 2023, and a month later, Jacksonville Beach Police Officer Scott O’Connell called Dillon and repeatedly accused him of the crime. Dillon said he was more than 300 miles away at the time and remembered thinking, “AI says I did this, how am I going to prove that I didn’t?”

The system used in Dillon’s case was known as FACESNXT, the Face Analysis, Comparison, and Examination System. According to CBS News and the ACLU lawsuit, police used cellphone photos taken from a computer screen showing surveillance footage, and FACESNXT returned a 93% facial-features match to Dillon. Dillon said he told police he had distinctive scars from skin cancer surgery running from his hairline toward his nose. After seeing the suspect images beside his own photo, he said the scars were “nowhere near alike.”

Dillon was arrested at home about eight months after that call. He spent a night in jail, borrowed money, and pledged the title to his truck to post bond. The charges were dropped about two months later, but the ACLU’s lawsuit says police left out evidence that could have weakened the warrant, including that a McDonald’s employee described the suspect as a regular, and license plate reader data showed no hits for Dillon’s vehicle near the restaurant.

The ACLU wrote in April that Kimberlee Williams, an Oklahoma grandmother, became the 14th publicly known person in the U.S. wrongfully arrested after police relied on facial recognition technology. That count did not include Richardson’s case, which surfaced later. Williams spent six months in jail after Maryland police pursued warrants tied to bank fraud. She said she could not have committed the crime because she was in Oklahoma. The ACLU also noted that more than 20 cities and jurisdictions have banned police use of facial recognition.

“When facial recognition technology generates false matches to innocent lookalikes, it can taint the investigation by tricking witnesses and police into mistakenly believing they’ve found the suspect,” the ACLU wrote.

In Richardson’s case, the JSO defended its work in a statement to Action News Jax. JSO said facial recognition was only one investigative tool, and that both the victim and the victim’s brother identified Richardson in photo lineups. The agency also said a judge signed a probable cause warrant, a second judge later found probable cause after a hearing, and the State Attorney’s Office reviewed the case before prosecutors dropped it. “Facial recognition technology is used as one tool among many available to investigators,” JSO said. “In this case, it was one tool, but certainly not the only tool.”

For the wrongfully accused, the psychological and emotional damage could last for years. “It’s ruined my life as far as being able to interact with children,” Dillon told CBS News. “I feel like people are watching me. I feel like people are saying: ‘hey, there’s that guy that was on the news, stay away from him.'”

Explore more on these topics

Share this story

Sign up for our newsletters

Subscribe and interact with our community, get up to date with our customised Newsletters and much more.