Man Sues Florida Cops Over 93% Facial Recognition Match

93% facial recognition match sparks federal lawsuit

A 93% facial recognition match triggered Robert Dillon's arrest. But he's now suing Florida law enforcement, alleging a botched investigation and withheld evidence in a 2024 incident where authorities used automated software to identify a suspect accused of attempting to lure a child at a Jacksonville Beach McDonald's. It's a shocking case. Dillon lived over 300 miles away in Fort Myers, yet he found himself caught in a criminal process that resulted in his arrest, jail time, and months of public scrutiny before charges were eventually dropped.

The mechanics of an automated error

Jacksonville Beach Police used FACES to generate the lead. This centralized database, maintained by the Pinellas County Sheriff's Office, holds millions of mugshots and driver license photos, and investigators fed a low-quality image into the system that produced the 93% facial recognition match score. But it's just a tool.

Officers then bypassed standard investigative steps that could have cleared Dillon immediately. The following investigative avenues were ignored despite being readily available:

• Mobile ordering records and payment data from the restaurant
• Cell phone location logs for the dates of the incident
• A physical comparison between Dillon and the suspect, specifically regarding a visible facial scar
• Time stamped surveillance footage and financial records

Concealing evidence from the court

The lawsuit details how officers secured an arrest warrant while omitting information that would have likely prevented charges, but they left key facts out of the affidavit. Corporal Scott O'Connell requested a search of license plate reader databases to track Dillon's vehicles. It returned zero results. So this negative result was left out, and O'Connell had previously spoken with Dillon by phone, during which the suspect described a distinctive scar and denied being in the area,that conversation was also absent from the affidavit.

This case is about what happens when police let an error-prone artificial intelligence system stand in for an investigation," the lawsuit said.

Dillon is represented by the American Civil Liberties Union and the Hoguet Newman Regal & Kenney law firm. But they assert the department promoted O'Connell even after his history of professional conduct issues. It's a troubling record. That history includes prior terminations and criminal charges from other jurisdictions that cast a long shadow over the promotion's justification.

Living with the fallout

The consequences hit Robert Dillon immediately. They lasted for years. He was arrested at his home while his family watched, held overnight in jail, and forced to pledge the title of his truck as collateral just to secure his release. But his mugshot remains online today. It creates a permanent record of an accusation that the state eventually abandoned.

A flawed identification process

What comes next

Dillon is seeking financial damages for the harm caused to his reputation and his livelihood. But the litigation aims to force substantive changes in how local law enforcement agencies utilize automated identification databases. This case highlights the tension between modern policing tools and the requirement for actual evidence by challenging the reliance on a 93% facial recognition match as sufficient grounds for an arrest. So the legal battle continues. For now, the parties in the US District Court for the Middle District of Florida await further proceedings.