This is an important decision in the fight for data protection. Monday, May 9, after two years of legal battle against the American Civil Liberties Union (ACLU), Clearview AI accepted stop selling its biometric database to private companies. The settlement has not yet been validated by a Chicago County judge.
It also predicts that the State of Illinois – in which the company was attacked – and its police will no longer be able to access it, for five years. On the other hand, federal services such as Customs and Border Protection, as well as law enforcement agencies in other states, will still be able to do so.
The association criticized Clearview AI, a company specializing in facial recognition, for failing to comply with Illinois laws on protecting sensitive data, in particular the Biometric Information Privacy Act, which prohibits companies from sharing certain data, such as a face or fingerprints. , without the consent of the interested party.
Clearview AI will also need to maintain the opt-out form that allows Illinois residents to opt out of photo search regardless of the affected customer. One operation, including communication, cost $50,000 in the deal. The form will also be available in California, where similar data protection regulations apply.
“This agreement shows that strong privacy laws can provide real protections against abuse.welcomed Nathan Freed Wessler, one of the deputy directors of the ACLU. Congress must act, and until they can, more states must compensate. »
“It’s a big win” , said Linda Xóchitl Tortolero, president of the Latin Women in Action association, which helps victims of domestic violence. In addition to the issue of consent, she points out the danger of ClearView AI’s collection of photos and other personal data, which can be used for bad purposes, especially by stalkers or even ex-spouses.
However, the company argues that it is a service that allows “make communities safer and enable law enforcement to solve crimes”. In the hunting list proudly displayed by the company, the 2020 capture of an Argentine child criminal: thanks to a blurred photo of the criminal, FBI investigators were able to find other photos of the latter, before finding his Facebook account and finally arresting him. It also helped identify protesters during the Capitol riot in January 2021.
Other countries are following suit
To build its database of 20 billion photos, Clearview AI scans social networks like Facebook or Instagram, and more generally the Internet, to retrieve all available photos. According to information from Washington post, the company recently raised ten million dollars to increase its activities. With its investors, it would have claimed to work to extend its database to 100 billion photos in a “that almost everyone on earth is identifiable”. A gigantic archive in which many citizens find themselves, without even knowing it.
Today, it is the legality of this process that the State of Illinois is questioning, and it is not alone. Canada, Italy, the UK and France are trying to take steps to protect their citizens. On December 16, the National Commission for Computing and Liberties (CNIL) notified Clearview AI to delete data collected in France and terminate its activities in the territory. She judged that the company had not ” a legitimate interest in collecting and using such data, in particular with regard to the particularly intrusive and massive nature of the process”. The American company replied that it had not “establishment in France or the European Union. It has no customers and does not engage in any activity that would otherwise mean it would be subject to [règlement général sur la protection des données] GDPR ».
In this battle for data protection, Clearview AI is not the first to have to give in to the judges. In 2020, Facebook, then attacked in Illinois for its facial recognition system, agreed to pay $650 million.