Florida man blames wrongful arrest on "error-prone" AI facial recognition

When police arrested Richard Dillon in 2023 for allegedly trying to lure a child away from a McDonald's in Jacksonville Beach, Florida, he told them he was more than 300 miles away at the time of the crime. The key evidence police used to puncture his alibi: facial recognition software matched an image of the suspect to Dillon's photo.

Dillon was later cleared. On Wednesday, he became a plaintiff in a new ACLU lawsuit filed against the Jacksonville Beach Police Department and others over what he believes was a case of misuse of AI-driven image matching technology.

"Police let an error-prone artificial intelligence system stand in for an investigation," the complaint argues. The case is the latest attempt to establish guardrails for powerful new technology that police are increasingly using to solve one of the toughest aspects of any investigation — when they have an image of a suspect, but not that person's identity.

Facial recognition is now an increasingly common law enforcement tool, with public databases holding images of 117 million Americans, according to the Center on Privacy and Technology at Georgetown Law School. The Dillon case illustrates what can go wrong when that technology is deployed without adequate safeguards.

The Arrest That Shouldn't Have Happened

The episode began in November 2023, when police say a man approached a 12-year-old in a McDonald's and tried to lure her away from her parents. A month later, Dillon received a call from Jacksonville Beach Police Officer Scott O'Connell, who accused him of the crime.

Dillon told CBS News he remembers thinking: "My life is over. AI says I did this, how am I going to prove that I didn't?"

During the call, Dillon denied involvement and described distinctive scars from skin cancer surgery that ran from his hairline down to his nose — physical features the suspect in surveillance footage did not have. It wasn't until Dillon was able to show evidence of those scars that the case against him unraveled.

By then, the damage was done. Dillon had been publicly identified as a suspect in a case involving a child. The emotional and reputational toll of being wrongfully accused of such a crime, even briefly, is nearly impossible to undo.

The ACLU's Legal Theory

The lawsuit argues that facial recognition technology was used as a substitute for proper police investigation rather than as one tool among many. The complaint contends that officers relied on the AI match as definitive identification without conducting the follow-up verification that responsible use of the technology requires.

This legal theory is significant because it doesn't challenge facial recognition itself — it challenges how police departments use it. The argument is that treating an AI-generated match as conclusive identification, without corroborating evidence, constitutes a deprivation of civil rights.

The case could set important precedent. If the court agrees that police must conduct independent verification of facial recognition matches before making arrests, it would establish a baseline standard that many departments currently lack.

A System With Known Flaws

The Dillon case isn't an isolated incident. Studies have consistently shown that facial recognition systems produce higher error rates for people of color, women, and older adults. A 2019 federal study found that some algorithms were 10 to 100 times more likely to misidentify Black and Asian faces compared to white faces.

Multiple wrongful arrests linked to facial recognition have been documented in recent years, including cases in Detroit, New Jersey, and Louisiana. In most of those cases, the people wrongly identified were Black men — a pattern that has drawn scrutiny from civil rights organizations and prompted several cities to ban police use of the technology outright.

What makes the Florida case particularly concerning is that the verification gap wasn't subtle. Dillon had visible scars that the suspect didn't have. A competent investigation would have caught the mismatch before it escalated to an arrest. The fact that it didn't suggests that the facial recognition result was treated as sufficient on its own.

The Regulatory Vacuum

There is no federal law governing how police departments use facial recognition technology. The FDA doesn't regulate it. Congress hasn't passed legislation establishing standards. And the patchwork of state and local regulations means that practices vary wildly from one jurisdiction to the next.

Some departments require that facial recognition matches be treated only as investigative leads, not as probable cause for arrest. Others, apparently including Jacksonville Beach, have used AI matches as the basis for direct confrontations with suspects — exactly the kind of substitution the ACLU lawsuit challenges.

The industry argument has always been that facial recognition is a tool, not a conclusion. But the gap between that stated position and actual police practice keeps widening, as the Dillon case demonstrates.

What This Means For You

If you think facial recognition is only a concern for people who look like criminal suspects, think again. With 117 million American faces in public databases, the odds of a false match affecting someone you know are not trivial. The technology is spreading faster than the rules governing it, and until courts or legislatures establish clear standards for verification and accountability, every false arrest is a preview of what could happen to anyone. Support legislation requiring independent verification before facial recognition results can be used as the basis for arrest. The technology can be useful — but only when it's a starting point for investigation, not a substitute for one.