The letter comes a day after the National Academies released a
106-page report concluding that facial recognition systems have gotten so advanced that the industry requires federal oversight. In the report, a committee of 14 experts heard from stakeholders on facial recognition over nine months, and concluded that the technology is a cause for concern because of both human misuse and technical shortcomings.
The Federal Trade Commission is actively policing the technology using existing consumer protection laws. The agency
issued guidance on the use of consumers’ biometric data last May, and in December, it issued a
first-of-its-kind ban on a private company using facial recognition, after alleging Rite Aid’s facial recognition system disproportionately misidentified people of color and women as shoplifters.
Facial recognition — the use of automated systems to identify people from surveillance footage, online photos or other images — has been on Congress’s radar for more than a decade. But the technology’s rapid advancement and adoption among law enforcement agencies and private industries in recent years is creating a new sense of urgency — as is new evidence that the technology can be harmful, ineffective or biased.
“The issue is more urgent than it has been in the past,” said Edward Felten, a former White House deputy U.S. chief technology officer under President Barack Obama, who helped write the National Academies report.
Thursday’s letter details instances where the technology misidentified people, resulting in false arrests. All six incidents in the letter affected Black people. One of the known false arrests is believed to involve Clearview AI’s facial recognition software,
according to a November letter from Sen. Ed Markey (D-Mass.) to the company.
Clearview AI’s CEO Hoan Ton-That defended his company’s software, noting that false arrests don’t stem from just the facial recognition mismatches.
“One false arrest is one too many, and we have tremendous empathy for the person who was wrongfully accused,” he said in a statement. “Even if Clearview AI came up with the initial result, that is the beginning of the investigation by law enforcement to determine, based on other factors, whether the correct person has been identified.”
The senator’s letter
also cited a POLITICO investigation finding that New Orleans police disproportionately used facial recognition on Black suspects.
The committee behind the National Academies report raised similar concerns with racial bias and facial recognition, noting that even if the technology were perfect, human bias could still create disparities.
Facial recognition technology “raises questions and introduces problems we associate with human decisionmaking, not just technical problems,” Elizabeth Joh, a law professor at UC Davis, and co-chair on the report’s committee, said.
Similar letters to the Justice Department in the past have highlighted the potential for racial bias with facial recognition, but little has happened in Congress despite these concerns.
An FTC commissioner noted that gap in December: “I urge legislators who want to see greater protections against biometric surveillance to write those protections into legislation and enact them into law,” Commissioner Alvaro Bedoya said in a statement on the Rite-Aid settlement.
Warnock’s office said the lawmakers want to develop a framework to regulate the growing use of facial recognition.
In addition to legislation, the National Academies, which often advises the government on scientific issues, used its report this week to call on the Biden administration to sign an executive order developing guidelines for how the government uses the technology.
“It is crucial that governments make tackling these issues a priority,” said Jennifer Mnookin, the chancellor of the University of Wisconsin-Madison and a co-chair on the committee that wrote the report, in a statement. Otherwise, she said Washington would “effectively cede” policy on a key public issue to private companies.
There are no federal regulations specifically on facial recognition in the U.S., though several laws have been introduced. Multiple lawmakers have proposed potential regulations for facial recognition since 2016, though all of them have fizzled out due to a lack of interest or conflicting opinions on how the technology should be regulated.
Artificial intelligence’s recent advancements, along with growing evidence of the effects of facial recognition’s racial bias, have led to renewed calls for regulations. But it will be difficult to pass any legislation with Republicans missing from much of the dialogue.
Even the FTC’s enforcement came from a panel of only Democrat FTC commissioners, while no Republicans signed onto Warnock and Durbin’s letter.
Facial recognition was once a bipartisan concern among lawmakers — with Republican lawmakers like Sen. Mike Lee (R-Utah) introducing
legislation limiting police use and then-House Rep. Mark Meadows
raising concerns about the technology, both in 2019.
But legislative efforts to regulate facial recognition have been led by Democrats and in states during the past four years. The lack of bipartisan participation could make it difficult for Congress to act, according to the National Academies.
“The issue is more urgent than it has been in the past. The urgency is increasing,” Felten said. “The committee has recommended that Congress consider legislation in this area. That said, legislation is difficult and Congress has many different things to do.”