U.S. lawmakers call for investigation into use and abuse of face recognition tech

Aug. 2, 2018

Less than a week after a damaging report exposed substantial flaws in facial recognition technology marketed to law enforcement by Amazon, five Democratic lawmakers are calling for an investigation into the commercial and government use, and potential abuse, of the technology.

In a letter, lawmakers raised concerns about the use of facial recognition and its impact on privacy rights, underscoring, in particular, the “disparate treatment of minority and immigrant communities within the United States.”

“Given the recent advances in commercial facial recognition technology—and its expanded use by state, local, and federal law enforcement, particularly the FBI and Immigration and Customs and Enforcement—we ask that you investigate and evaluate the facial recognition industry and its government use,” the lawmakers wrote in a letter to Gene Dodaro, head of the U.S. Government Accountability Office (GAO). The congressional watchdog’s prior research into facial recognition, they said, has been instrumental in educating the public and Congress “as to the use, and possible misuse, of these technologies.”

The letter was signed by Senators Ron Wyden, Chris Coons, Ed Markey, and Corey Booker, and Jerrold Nadler, the ranking Democrat on the House Judiciary Committee.

In a list of topics, lawmakers asked the GAO to investigate specifically which law enforcement agencies are using the technology and how. It’s unclear, for example, which agencies have acquired Amazon’s facial recognition technology, known as Rekognition; the company’s client page only identifies one local law enforcement agency by name.

The lawmakers further asked the GAO to evaluate what types of safeguards are in place, if any, to prevent government agencies from violating the constitutional rights and personal privacy of individuals subjected to the technology.

Moreover, the GAO was asked to consider whether police are deploying the technology in areas with prior complaints of discriminatory policing, such as low-income or minority communities, and whether that might “exacerbate discrimination and other policing disparities.” The lawmakers asked that those tasked with evaluating facial recognition actually purchase the commercial products available to test them, and investigate whether the programs have a “disparate impact on certain racial or ethnic groups.” (Studies have repeatedly shown facial recognition to be far less accurate when used on people with darker skin.)

There is also growing concern about the potential sale of data gathered via commercial facial recognition software, which is widely in use but often without the knowledge or consent of the public. It was revealed recently, for instance, that a Canadian shopping mall was using facial recognition to secretly track shoppers. The mall, Chinook Centre in Calgary, said it didn’t require consent because the software—which was revealed to the public by mistake—doesn’t capture or retain facial images.

The American Civil Liberties Union said it had conducted tests using Amazon’s Rekognition software using 25,000 publicly available mugshots and photos of all 535 members of Congress. According to the ACLU, Rekognition misidentified 28 members of Congress, listing them as a “match” for someone in the mugshot photos. The error rate was significantly higher for non-white members of Congress. Amazon noted in response that its default confidence threshold is 80%, but that when “highly accurate face similarity matches are important,” it recommends a setting of 99%.

Whether police equipped with Rekognition are using a 99% -threshold setting—and whether that’s even likely to solve the technology’s inherent racial bias—is unknown. But errors in identifying suspects could have deadly consequences, as the ACLU of Northern California points out: “If law enforcement is using Amazon Rekognition, it’s not hard to imagine a police officer getting a ‘match’ indicating that a person has a previous concealed-weapon arrest, biasing the officer before an encounter even begins.”

In a letter, Senator Markey and Congressmen Luis Gutiérrez and Mark DeSaulnier asked Amazon to address many of the same questions the GAO is being asked to investigate now. “While facial recognition services might provide a valuable law enforcement tool, the efficacy and impact of the technology are not yet fully understood,” the lawmakers wrote in a letter to Amazon CEO Jeff Bezos. “In particular, serious concerns have been raised about the dangers facial recognition can pose to privacy and civil rights, especially when it is used as a tool of government surveillance, as well as the accuracy of the technology and its disproportionate impact on communities of color.” (Bezos has not yet responded to the letter.)

Gizmodo has the full story

Sponsored Recommendations

Six Cloud Strategies to Combat Healthcare's Workforce Crisis

The healthcare workforce shortage is a complex challenge, but cloud communications offer powerful solutions to address it. These technologies go beyond filling gaps—they are transformin...

Transforming Healthcare with AI Powered Solutions

AI-powered solutions are revolutionizing healthcare by enhancing diagnostics, patient monitoring, and operational efficiency - learn how to integrate these innovations into your...

Enhancing Healthcare Through Strategic IT and AI Innovations

Learn how strategic IT and AI innovations are transforming healthcare - join Tomas Gregorio as he explores practical applications that enhance clinical decision-making, optimize...

The Intersection of Healthcare Compliance and Security in the Age of Deepfakes

As healthcare regulations struggle to keep up with rapid advancements in AI-driven threats like deepfakes, the security gaps have never been more concerning.