Who in America is resisting privacy-disrupting facial recognition technology? Well, let’s start with Maine … • The Registry

Functionality The state of Maine enacted what the American Civil Liberties Union (ACLU) describes as the strictest state facial recognition law in the United States amid growing concern over the unconstrained use of the systems facial recognition by the public and private sectors.

The Maine Bill, LD 1585 [PDF], prohibits state officials from using facial recognition technology or entering into agreements with third parties to do so, except in a relatively limited set of circumstances related to serious crimes and license data searches vehicles. It imposes the kind of blanket limitations that civil liberties groups advocate.

“Maine is showing the rest of the country what it looks like when we the people control our civil rights and freedoms, not the tech companies that could profit from the government’s widespread use of facial surveillance technology,” said Michael Kebede, head of policy. lawyer at the ACLU in Maine, A declaration.

“The overwhelming support for this law shows the Mainers agree that we cannot let technology or tech companies dictate the contours of our basic constitutional rights.”

Meanwhile, on Tuesday, the United States Government Accountability Office (GAO) released a report Recommend that U.S. federal agencies take steps to better understand and manage how they use facial recognition technology. Twenty federal agencies currently use facial recognition, whether it’s their systems or third-party systems like Clearview AI or Amazon Rekognition. Fourteen rely on non-federal systems, but 13 of them were unable to provide details of what their employees were using at GAO.

And two weeks ago a set of Democratic lawmakers reintroduced the Moratorium Law on Facial Recognition and Biometric Technology, in response to reports of unregulated use of such systems.

But not all laws are so strict. The ACLU points to the Washington Facial Recognition Bill (SB 6280), passed in 2020 with the Microsoft support, as an example of weak legislation that does not protect against mass surveillance.

Critics say the technology is unreliable, discriminatory and violates civil rights. For example, in January 2020, Robert Julian-Borchak Williams was mistakenly arrested in Michigan following a faulty facial recognition match.

Meanwhile, trade gets involved

Advocacy groups have also sounded the alarm bells about the inconspicuous use of facial recognition technology in commercial applications. Fight for the future, a digital rights group, highlights the use of facial recognition by the Stripe payment system and the Tinder hangout as examples of the trend.

Strip Identity, announced two weeks ago, provides a service that allows businesses to identify customers by verifying government identities and selfies. “Stripe’s identity verification technology uses computer vision to create temporary biometric identifiers of your face from selfies and your photo ID photo, and compares the two,” the biz Explain.

Stripe claims that it does not see biometric identifiers in its system and deletes those identifiers after 48 hours.

Tinder also uses facial recognition for its photo verification service at least since june 2020. He recently added a pop-up in his app revealing the use of facial recognition technology.

The register asked if this had been done for a particular reason. A spokesperson for Tinder said, “No, we just encourage members to verify their profile with photo.”

Click to enlarge

Caitlin Seeley George, Director of Campaigns and Operations for Fight for the Future, said The register by email that facial recognition should be banned altogether.

“We need a federal ban on facial recognition,” she said. “This technology is too dangerous to be regulated, it should not be used at all.”

The use of facial recognition by Stripe and Tinder, she said, is of deep concern.

“These facial recognition systems will certainly lead to discrimination: not only will there be cases where the technology cannot correctly identify users, but companies could also use this tool to actively prohibit certain people from buying their products or from being able to use them. ‘use the apps,’ explained Seeley George.

“They are also normalizing the use of facial recognition in people’s lives and further accelerating its adoption in society in a way that violates the rights of people and puts them at risk.”

When asked if there is value in facial recognition, Seeley George said the technology is doing more harm than good.

“It doesn’t work, it creates biometric databases that can be targeted by those who want to do wrong, and it opens up more possibilities for supercharged discrimination,” she explained.

“When a company like Stripe makes facial recognition available to all users, the technology seems normal and acceptable to people, but this standardization is very problematic. People shouldn’t feel comfortable scanning their face every time they want to buy something or access content. online.”

A mixed approach

Another online advocacy group, the Electronic Frontier Foundation, would prefer to see more measured limits on facial recognition technology based on the fact that certain uses can be positive.

“It does not follow that all private use of facial recognition technology infringes human rights,” EFF said in a publication earlier this year. “For example, many people use facial recognition to lock their smartphones.”

A complete ban on the technology, the EFF argues, would also ban research and activist uses.

Rachel Thomas, co-founder of fast.ai and the founding director of the Center for Applied Data Ethics at the University of San Francisco, said The register by e-mail that she is also in favor of targeted regulation.

“My view is that we would need to ban particular applications of facial recognition (for example, ban its use in commercial or government contexts) rather than ban the technology in the abstract,” he said. she declared.

“I think the debates over whether to prevent anyone, in any setting, from doing facial recognition can be distracting, and that it is more useful to focus on practical applications and those potentially affected. . “

Given the number of tech companies around the world using facial recognition and the usefulness of the technology for law enforcement and commerce, it’s hard to believe that those with a financial interest in facial data would quietly accept laws banning facial recognition.

“Companies are already trying to rename facial recognition because the term has become so toxic,” said Seeley George.

“Microsoft recently started providing Australian police with ‘object recognition’ services, which you can imagine could be used to try to identify or track specific people. Other companies describe them. things like “face scans” instead of “facial recognition” to try to avoid scrutiny. So there is no doubt that companies will try to avoid being targeted by this legislation. “

Nonetheless, she argues that nothing less than a total ban will suffice.

“At the end of the day, facial recognition is dangerous no matter who uses it, and businesses and ruling private entities are already using the technology in ways that put people at risk,” she said. .

“Just as facial recognition strengthens racist policing, private entities can use facial recognition to discriminate against people and it can increase societal prejudice. “

“So yes, we have to ban everything. That said, passing federal facial recognition ban legislation would still be a very real step forward that will protect those targeted by the government and the use of the technology by them. law enforcement.” ®