All incidents

36-year-old Black man wrongfully arrested after facial recognition misidentification in New York leading to two days in custody

Apr 1, 2025New York, United States3 sources

Summary

Trevis Williams, 36, was wrongfully arrested in April in New York after a facial recognition system incorrectly matched his mug shot to a suspect in a Manhattan flashing case. Despite physical and location discrepancies, Williams was arrested and spent over two days in custody before the case was dropped. Facial recognition technology, which converts faces into data points for comparison, has well-documented racial biases, with error rates significantly higher for Black and Asian faces compared to white ones. The NYPD uses facial recognition regularly, though it is supposed to serve only as an investigative lead, not as sole evidence for arrest. At least 10 similar wrongful arrests linked to facial recognition have been reported nationwide, prompting calls from civil rights groups for stricter oversight and transparency in its use. Williams’s case highlights the risks of combining flawed algorithms with unreliable eyewitness testimony, particularly impacting minority communities overrepresented in police databases.

Incident Details

Domain
Algorithmic Discrimination

Automated systems that produce discriminatory outcomes based on protected characteristics.

Harm Types
Wrongful Arrest
Discrimination

Differential treatment or outcomes based on protected characteristics.

Mechanism
conduct
Recipient
IndividualTrevis Williams
Dimensions
autonomyreputationalpsychologicaldiscriminatory

Who Was Affected

Age
Adult
Gender
Male
Group
Racial/Ethnic Minority