by Micaela Burrow
The Department of Defense (DOD) and the FBI collaborated on an artificial intelligence-driven facial recognition technology program provided to at least six federal agencies and a Pentagon agency that supports civilian police forces, The Washington Post reported.
The facial recognition software could be used to identify individuals whose features were captured by drones and CCTV cameras, the Post reported, citing documents obtained through a Freedom of Information Act request as part of an ongoing lawsuit by the American Civil Liberties Union (ACLU) filed against the FBI. The documents reveal federal authorities were more deeply involved in development of the technology than was previously known, sparking concerns over Americans’ privacy rights.
“Americans’ ability to navigate our communities without constant tracking and surveillance is being chipped away at an alarming pace,” Democratic Sen. Ed Markey of Massachusetts told the Post. “We cannot stand by as the tentacles of the surveillance state dig deeper into our private lives, treating every one of us like suspects in an unbridled investigation that undermines our rights and freedom.”
One example is the Janus program, a project out of the U.S. intelligence community’s research arm, known as Intelligence Advanced Research Projects Agency (IARPA), the documents show, according to the Post. JANUS worked on software to harness “truly unconstrained face imagery” obtained through public surveillance cameras.
The IARPA program manager said in 2019 the software could “dramatically improve” facial recognition with “scaling to support millions of subjects” and identify faces from obstructed distances. One version could identify faces “at target distances” of more than a half-mile.
Research teams constructed new algorithms aimed at “radically expanding the scenarios in which automated face recognition can establish identity,” the documents show, according to the Post. Some emails feature FBI employees discussing how the software could process images using attributes “face rectangle x start coordinate,” “pitch of the head” and “probability of being male” with researchers and engineers.
In 2017, DOD paid thousands of volunteers to test the system at distance and in a variety of simulated scenarios, including at a hospital, a subway station, an outdoor marketplace and a school, the documents show, according to the Post.
The final version of the software, nicknamed Horus, was provided to the Combating Terrorism Technical Support Office within the DOD, according to the Post. The counter-terrorism office shares military technologies with civilian police forces
At least six federal agencies operate Horus, providing feedback to DOD that then was applied “to refine the tool,” according to the Department of Homeland Security as of May 2022.
Though today’s #AI technology is useful, it lacks certain capabilities to keep our nation and allies safe. Last week, we announced AI Forward, our upcoming initiative for reimagining the future of AI for national security. More at: https://t.co/hF9BjXhEJq pic.twitter.com/KJkhcrDZDf
— DARPA (@DARPA) October 31, 2022
Three states and at least a dozen cities have passed laws banning or restricting their police forces from employing facial recognition technology amid concerns over civil liberty violations, according to the Post.
The FBI and Department of Defense did not provide comments to the Post.
– – –
Micaela Burrow is a reporter at Daily Caller News Foundation.
Photo “FBI Agent” by FBI – Federal Bureau of Investigation. Background Photo “U.S. Pentagon” by U.S. Secretary of Defense. CC BY 2.0.
