Amnesty International Calls for Ban on Police Use of Facial Recognition

International NGO Amnesty International is the latest organization to call for a ban on the use of facial recognition technology by law enforcement agencies.

The announcement comes on the heels of a recent investigation published by Amnesty International where it states that the NYPD has more than 15,000 cameras that employ “invasive and discriminatory facial recognition software” in the boroughs of Manhattan, Brooklyn, and the Bronx.

Facial recognition technology has been under intense scrutiny following a number of high-profile stories that shed light on some of its more nefarious uses.

In January of 2020, a front page story in The New York Times revealed that startup Clearview AI had scraped public images from social media sites such as YouTube, Facebook, and Twitter to develop its facial recognition algorithm which it then began selling to law enforcement agencies throughout the world.

In China, a number of companies with close ties to the government have been accused of using facial recognition to monitor and oppress the Uighur Muslim minority living in the country’s far Northwest Xinjiang region, leading to the blacklisting of several public and private entities that have been linked to the profiling of Uighurs by the United States government.

In this most recent case, Amnesty’s investigation reports that the NYPD has used facial recognition technology in roughly 22,000 cases since 2017, with half of those occurring in 2019 alone.

In a press release announcing the study and declaring its condemnation of the NYPD’s use of the technology, Amnesty notes that in one area of Brooklyn with a predominantly Black (54.4 percent) and Hispanic (30 percent) population there were 577 cameras found at intersections, making it the most surveilled neighbourhood in all three boroughs.

The notion that facial recognition technology exhibits racial bias is not new, and has accelerated since a December 2019 NIST study revealed that a number of popular facial recognition algorithms are far more likely to misidentify people of color (especially Black women) as opposed to white men.

Source: Al Jazeera

–

June 10, 2021 — by Tony Bitzionis