A group of over twenty artificial intelligence (AI) researchers within the tech and academic world has asked Amazon Web Services (AWS) to stop selling the facial recognition software Recognition to the police.
Critics of police use of rekognition include Yoshua Bengio, who recently won the Turing Award, as well as former Amazon principal scientist and Caltech professor Anima Anandkumar. The group also includes researchers in computer vision and machine learning from Google AI, Microsoft Research and Facebook AI Research, writes Venturebeat.
Rekognition has been used by police in Florida and Washington and would also have been offered to the U.S. Department of Homeland Security to identify immigrants. “We call on Amazon to stop selling to Rekognition to the police as there are no rules and restrictions to prevent abuse,” said the researchers in an open letter.
“The researchers also quote work by privacy advocates who are concerned that police and other organizations with little understanding of the technical aspects of computer vision systems may make serious mistakes. Think of arresting and imprisoning an innocent person.
In addition, the research criticizes Recognition for its binary classification of sexual orientation as male or female, an approach that can lead to misclassifications.
Criticism has also been levelled at statements by deep learning and AI-general manager Mathew Wood and global head of public policy Michael Punke, who reject the results of a recent audit, which showed that the software wrongly identifies dark skin women as men in 31 percent of cases.
The researchers’ public letter also comes a week after members of the American Senate proposed a law to regulate the use of facial recognition software. Amazon then said that it welcomed a form of regulation or “legal framework”. Microsoft urged the federal government to regulate facial recognition software before police and other organizations abuse it.This news article was automatically translated from Dutch to give Techzine.eu a head start. All news articles after September 1, 2019 are written in native English and NOT translated. All our background stories are written in native English as well. For more information read our launch article.