New documents revealed by the ACLU and shared with Gizmodo show the lengths FBI and Pentagon officials went to develop “truly unconstrained” facial recognition capable of being deployed in public street cameras, mobile drones, and cops’ body cameras.
The goal of the project, code-named “Janus” after the Roman god with two opposing faces, was to develop highly advanced facial scanning tech capable of scanning people’s faces across a vast swath of public places, from subway cars and street corners to hospitals and schools. In some cases, researchers believed the advanced tech could detect targets from up to 1,000 meters away.
Experts speaking with Gizmodo said the advanced surveillance capabilities outlined in the documents potentially pose “truly unprecedented threats” to personal privacy and civil liberties, especially given the U.S.’s lack of any meaningful federal privacy protections. If implemented as documented, the Janus program’s blanket of cameras would resemble surveillance systems already in place in China, which the Pentagon and other intelligence agencies have publicly denounced.
“The government is opening Pandora’s box on a terrifying technical capability, which can enable pervasive tracking of anyone or everyone in a way never before possible in a free society,” ACLU Speech, Privacy, and Technology Project Deputy Director Nathan Freed Wessler told Gizmodo. “Lawmakers need to close the door on government abuse of this technology now, before it is too late.”
What is the Janus Program?
The documents, first reported on by The Washington Post, provide new insights into the nearly decade-old surveillance research program funded by the Intelligence Advanced Research Projects Agency, (IARPA), a U.S intelligence group modeled after DARPA, the Pentagon’s research and development unit. The Janus program dates back to 2014, when it was launched with the stated goal of “radically expanding the scenarios in which automated face recognition can establish identity.”
IARPA researchers cited in the documents said they were interested in dramatically improving the quality of facial recognition systems and allowing, “scaling to support millions of subjects.” The documents reveal the researchers were interested in creating scanning tech capable of quickly detecting faces from partially obstructed angles from a distance of more than half a mile away. One of the documents says image data was collected from a camera on a “small fixed-wing unmanned aerial vehicle” flown over a marketplace. The researchers’ vision of the technology as deceived in the documents paints a picture eerily similar to the type of always-on, high-powered, public surveillance system currently in use in China and Russia.
FBI agents and academics working on the project appeared acutely aware of the power and surveillance potential of modern devices increasingly entrenching themselves deeper in people’s lives. One FBI scientist, according to the Post report, referred to cell phones and social media as some of the “biggest enablers of better face recognition.”
Speaking with Gizmodo, Surveillance Technology Oversight Project Executive Director Albert Fox Cahn said all forms of facial recognition pose privacy issues but said the scope of government surveillance capabilities explored by the Janus program raises “profound concerns.”
“This system would pose truly unprecedented threats,” Fox Cahn said. “With wide-area facial recognition systems, intelligence agencies could track our movements across entire cities with a small number of cameras.”
Federal agents, according to one of the academic researchers working on the project speaking with the Post, tried to draw distinctions between surveillance capabilities they intended to use domestically and those they wanted to deploy in other parts of the world. However, as debates over NSA surveillance tools unearthed by whistleblower Edward Snowden show, those once-clear lines become incredibly opaque once a powerful technology is deployed.
“The question always in the back of my mind was: What does the intelligence community really want to do with this stuff?” Erik Learned-Mille, a University of Massachusetts at Amherst involved with the research said in an interview with the Post.
The DoD and FBI did not respond to Gizmodo’s request for comment.
Janus officially ended in 2020, however, as the Post notes, its work was included in a web-based face search tool called Horus provided to the DoD’s Combating Terrorism Technical Support Office. Meg Foster, a Justice Fellow at Georgetown Law’s Center on Privacy & Technology told Gizmodo the fact the program has ended provides little comfort.
“While Janus may have ended, these documents demonstrate that the prevailing approach to surveillance technology is to put the cart before the horse, and given the growing number of wrongful arrests stemming from face recognition, we know that the most vulnerable among us will experience the consequences first and hardest,” Foster said
The documents were unearthed come in response to a Freedom of Informaiton Act lawsuit the ACLU filed against the FBI. The most recent documents reportedly date back to 2019 when the ACLU sued the government for their release
Facial recognition tech ‘ripe for abuse’
U.S. Facial recognition has absolutely exploded in recent years, with the tech being used to scan faces at sports stadiums and concert venus to airport terminals and iPhone lock screens. As of 2021, according to a Government Accountability Office audit, at least 20 federal agencies use facial recognition, though it’s unclear if those uses have any connection to the more powerful tools described in the ACLU documents.
Still, prolific as the tech may be, the vision outlined in the IARPA docs appears to describe something very different: a powerful, pervasive, and public surveillance apparatus with a wide net potentially capable of implicating everyday people simply trying to catch a subway to work or walk home. Foster warned a system with those capabilities deployed in U.S. cities risks creating a “perpetual lineup in which we are all suspects.”
“The Janus program is a deeply alarming confirmation of what privacy and civil rights advocates have long warned—that facial recognition technology is absolutely ripe for abuse,” Foster added.
States and cities show growing divide over facial recognition
The FBI and DoD’s attempts to move the goalpost on American surveillance systems counterintuitively come during a time of heightened scrutiny on the technology, particularly amongst local governors. To date, at least 16 municipalities, including San Francisco Boston, and Oakland already passed laws and ordinances on facial recognition use by law enforcement or in the public. Private tech giants Like Amazon and Microsoft have banned police use of facial their facial recognition tech, though loopholes may exist for some federal agencies.
States and cities may continue to move the needle on biometric privacy protections, but the same can’t be said for the federal government which still lacks any meaningful privacy protections. Fox Cahn, of the Surveillance Technology Oversight Project, said the Janus revelation provides even more urgent incentives to pass some sort of federal forms.
“This sort of surveillance is not only creepy but unconstitutional, amounting to the type of persistent location tracking the Supreme Court has struck down with other technologies,” he said. “But we shouldn’t wait years for this type of tracking to be struck down as well. We need to enact comprehensive facial recognition bans today.”