Ousmane Bah filed a federal lawsuit in April 2019 against Apple seeking $1 billion in damages. The lawsuit claims that Bah was falsely accused of a series of thefts from Apple stores based on the company’s use of a facial recognition system to identify shoplifters. Bah was charged in four states for theft but once authorities realized that he was wrongly accused, charges were dropped in three cases with a likely dismissal in the fourth. The lawsuit contends that Bah suffered significant damage to his reputation because of a mistake based on the company’s use of facial recognition software.
Also in April, Amara Majeed, a Brown University student, was falsely identified as a suspect in a Sri Lanka bombing. Sri Lanka police issued a press release where Majeed’s photo was mistakenly placed next to the name of a suspect. People recognized Majeed from her Facebook profile page and called her a terrorist. After the police issued a retraction, Majeed continued to receive defamatory comments and threats.
Due partly to these accounts, Congress held hearings in May and June on facial recognition technology and civil liberties. At the hearings, testimony revealed that 18 states share photo images with the FBI with the result that over 50% of American adults are in a facial recognition database. Testimony also stated that databases for facial recognition are using images from social media sites where people are posting photos of themselves and others in astonishing numbers.
Experts at the hearings discussed a number of limitations of facial recognition technology. The technology is not very effective if the subject is moving, is part of a crowd, or a number of objects are in the image. Facial appearance can be altered due to age, weight fluctuation, change in facial hair, and cosmetic surgery leading to misidentifications. The technology is most effective for white males and makes more errors with minorities, particularly African Americans.
You have free articles remaining.
Despite these limitations, facial recognition technology has been applied in a number of ways in the private sector. Certain hotels at check-in use it to greet guests by name, and some casinos identify problem patrons with the technology. Bars use it to identify underage drinkers, and a matchmaking site employs the technology to link people who look alike. Stores use it to identify repeat customers and shoplifters. And Facebook uses the technology to identify people and tag them in other people’s photos for potentially all Facebook users to view.
Law enforcement has also applied the technology in several ways. At the 2001 Super Bowl, fans were scanned to identify potential criminals and terrorists. A few people were identified with minor criminal records but several were misidentified. More recently, the Baltimore police used facial recognition on people publicly protesting the death of Freddie Gray. Several had outstanding arrest warrants but the images of all were kept. The technology has located identity theft and fraud suspects. And, using videos, police identified assailants involved in fights. With the increase in body cams, police are likely taking and retaining more images of people.
Government has been slow to respond to the use of facial recognition technology and impose any regulations that would protect First and Fourth Amendment rights along with the right to privacy. Companies have yet to impose meaningful boundaries on sharing images for facial recognition databases or applications for the technology. So smile nicely for the next selfie that you post to a social media site because a foreign agent might use the image and your name to create a false profile to spread misinformation for the 2020 election, and you will want to look good.