A Florida man, Robert Dillon, sues over a wrongful arrest after faulty AI facial recognition misidentified him as the McDonald’s suspect in Jacksonville Beach, despite Dillon living 300 miles away in Fort Myers.

The suit spotlights how facial recognition in policing often yields candidates rather than definitive matches and can trigger wrongful arrests without thorough corroboration.

Investigators matched surveillance to a suspect photo and relied on a victim’s claim in a separate Jacksonville car-theft case, leading to Dillon’s arrest.

The case was later dropped after investigators confirmed the suspect did not commit the crime.

All charges against Dillon were dropped, but the experience caused ongoing personal and financial stress and lasting fear involving interactions with children.

Facial recognition was used as one tool among others; authorities say it contributed to probable cause, though prosecutors dropped the charges after a year-long review.

Dillon describes the experience as traumatic, seeking accountability and justice for the impact on his life.

The timeline: investigation opened in early April 2025 and was resolved by May 2025, underscoring concerns about AI-driven policing accuracy and accountability.

Vehicle data and license-plate searches failed to corroborate the arrest and were reportedly omitted from the warrant.

Authorities stated facial recognition was one tool used, not the sole basis for action, while Dillon works to rebuild his life and secure employment.

A police phone call repeatedly framed Dillon as the offender, despite his denials and explanations about scars from skin cancer surgery.

The case culminated in a year-long effort to drop the charges and remove the arrest record.