Man Wrongfully Accused by Facial Recognition

From Privacy Wiki
Jump to navigation Jump to search
Man Wrongfully Accused by Facial Recognition
Short Title Michigan Man Wrongfully Accused by Facial Recognition Algorithm
Location Michigan
Date June 2020

Taxonomy Aggregation, Identification, Distortion, Exclusion
Information Physical Characteristics, Identifying
Threat Actors Investigator of the case, DataWorks Plus, NEC, Rank One Computing, Law Enforcement

Individuals
Affected A wrongfully accused man
High Risk Groups Ethnic Minority
Secondary Consequences Anxiety, Incarceration

A Michigan man was wrongfully accused by a facial recognition algorithm, which led to his arrest.

Description

The Shinola store (upscale boutique in Detroit) shoplifting occurred in October 2018. An investigator reviewed the store’s surveillance video and sent a copy to the Detroit police. Five months later a digital image examiner for the Michigan State Police, uploaded a “probe image” — a still from the video — to the state’s facial recognition database. The system would have mapped the man’s face and searched for similar ones in a collection of 49 million photos. Aggregation

After police ran the search of the probe image, a wrong man’ driver’s license photo was among the matches. Identification

Faulty facial recognition match led to a Michigan man’s arrest for a crime he did not commit. Distortion

He had no way of knowing, that his photo was used by the algorithm in a search.Exclusion

In Michigan, the DataWorks Plus software is used by the state police and it incorporates components developed by the Japanese tech giant NEC and by Rank One Computing, based in Colorado.

National Institute of Standards and Technology (NIST) have found that while the technology works relatively well on white men, the results are less accurate for other demographics, in part because of a lack of diversity in the images used to develop the underlying databases.

In response to media coverage of the case, the Wayne County prosecutor’s office apologized and said that the wrongfully accused man could have the case and his fingerprint data expunged.

Risk Statistics

Laws and Regulations

Sources

https://www-nytimes-com.cdn.ampproject.org/c/s/www.nytimes.com/2020/06/24/technology/facial-recognition-arrest.amp.html