This week’s riots in the U.S. Capitol could herald a shift in attitudes towards the use of facial recognition surveillance by law enforcement, according to digital rights experts interviewed by Coda Story.
Many of the rioters who mobbed the U.S. Capitol on Wednesday did not cover their faces or use any of the elaborate schemes protesters worldwide have developed in recent years to avoid identification by police cameras — leading to calls on social media for facial recognition software to be used to track and arrest them.
While facial recognition is not known to have been used in the arrests of more than 80 Capitol protesters, a debunked, and since retracted, story in the Washington Times claimed facial recognition technology had been used to identify the presence of “Antifa members” in the crowd.
The protest in Washington occurred at a moment in which public perception of facial recognition has dramatically soured, with outright bans of police use of the technology in cities like San Francisco and Portland. Recent studies have also pointed to the disproportionate inaccuracy of automated facial recognition technology in recognizing darker-skinned people.
According to John Honovich, founder of the video surveillance industry publication IVPM, over the last year, “the sales of facial recognition and the interest in facial recognition have significantly declined directly in response to the ethical and political concerns of the technology. We’ve seen layoffs in numerous of the most notable public companies in the space.”
Steven Feldstein, a fellow at the Carnegie Endowment for International Peace who researches the global implementation of AI surveillance technology, said incidents like the Capitol riot could prompt a rethink about facial recognition. “As much as many of us have been criticizing overreach when it comes to the surveillance aspects associated with these technologies, you can also see in instances like this where it can be very helpful for identifying people committing criminal acts and violence.”
Feldstein added that the September 11, 2001 attacks provide a historical precedent for such a shift in attitudes. “For a long time, 9/11 really shifted the perspective of people who normally would err on the side of pushing for greater protections in civil liberties, protections for a variety of individuals. In the face of such an egregious act, I think a lot of people were willing to err on the side of providing law enforcement capabilities, whether it was enhanced surveillance authorizations or other related powers,” he said in a phone interview.
Digital rights advocates remain concerned about the effectiveness of facial recognition. Matthew Guariglia of the Electronic Frontier Foundation, a U.S.-based nonprofit organization which supports a full ban on government use of the technology, suggested facial recognition is unlikely to be very useful in identifying the D.C. rioters. “A lot of these people that law enforcement are searching for are very much already known, if not to law enforcement, then to people who use social media, to people who have been to protests in the past few years,” he said.
“But there’s always a temptation after a traumatic incident like the one we saw yesterday, to put all of our eggs in one basket, and that basket be more invasive surveillance technology,” he added.
Feldstein cautioned against making predictions too early. “While we’re still in the midst of figuring out what happened, it’s a little hard to know what the fallout will be,” he said. “But my sense is in a few different areas, there will be a push for potentially giving a little more credence to law enforcement to take the actions that their people believe are necessary to bring accountability in public order to a really damaging episode.”