Amazon technology has confused 28 congressmen with criminals



[ad_1]


Amazon facial recognition surveillance technology became the target of several national events . Now, 28 other compelling reasons arise because there is such concern.

The American Civil Liberties Union (ACLU) recently performed a tool test called " Rekognition " and surprisingly the software misunderstood. to 28 members of the North American Congress, identifying them as people who were arrested for a crime, informed the civil organization.

Members of Congress who were confused with the database of photographs that were used in the test include Democrats, men and women, legislators of all ages and from all over the country, said ACLU.

However, was surprised that recognition errors were mostly with people of color .

To perform the test, exactly the same facial recognition system that Amazon offered was used. public, that anyone could use to find matches between the images of faces. The use of Rekognition allowed to build a database and a search tool with the help of 25,000 photos of people who had been photographed with the help of Rekognition. Arrested and Available to the Public

They then searched the public photos of each current member of the House and Senate in this database. "We use the default match setting that Amazon sets for Rekognition," the ACLU said on its website.

In a recent letter to the CEO of Amazon Jeff Bezos The Congressional Black Caucus expressed concern about the "unforeseen and profound negative consequences" that surveillance could have for people of color, undocumented immigrants and protesters. Our results validate this concern: almost 40% of the false coincidences of Rekognition in our test came from people of color even if they represent only 20% of the Congress.

If the police are using Amazon Rekognition, it's not hard to imagine that a police officer will get a "coincidence" stating that someone has already been arrested with a concealed weapon, which predisposes the officer even before the start of a meeting. 19659002] An identification, accurate or not, could cost people their freedom or even their life . People of color have already been disproportionately harmed by police practices, and it's easy to see how Rekognition could exacerbate that.

A recent incident in San Francisco provides a disturbing illustration of this risk. The police stopped a car, handcuffed an old black woman and forced her to kneel under the threat of a rifle, all because a license plate reader. automatic registration incorrectly identified his car as a stolen vehicle. @amazon @ awscloud That's exactly why we wrote to you in May. https://t.co/vNyAnuhJae

– Black Caucus (@OfficialCBC) July 26, 2018

Associating people with pictures of people who have been arrested is not a hypothetical exercise . Amazon is aggressively marketing its facial surveillance technology to the police, baduming that its service can identify up to 100 faces in a single image, tracking people in real time through surveillance cameras and scanning body camera images .

The department of a sheriff in Oregon has already started using Amazon Rekognition to compare people's faces with a database of police photos, without any public debate.

These dangers explain why Amazon employees, shareholders, a coalition of nearly 70 civil rights groups, more than 400 members of the academic community and more than 150,000 members of the public have already Demonstrated to demand that Amazon stop providing facial surveillance to the government

Congress must take these threats seriously, curb and adopt a moratorium on the use of the law for facial recognition. This technology should not be used until the damage has been fully taken into account and all necessary measures have been taken to prevent harm to vulnerable groups.

[ad_2]
Source link