Responding to call from ethics board, leading producer of police body cameras agrees to keep face recognition off its products
June 27, 2019- Today, Axon, the leading manufacturer of police body-worn cameras, has agreed to keep face recognition technology from its products, responding to a call from its own independent ethics board.
In a report issued today, after a year of meetings and research, Axon’s AI and Policing Technology Ethics Board concluded that face recognition technology is not yet reliable enough to justify its use on body-worn cameras, and expressed particular concern regarding evidence of unequal and unreliable performance across races, ethnicities, genders and other identity groups. Given these findings, the Board called on Axon to commit to not proceed with the development of face matching products, including adding these capabilities to body-worn cameras.
Axon agreed to this and other recommendations of the Board.
Axon’s AI and Policing Technology Ethics Board operates independently from the company and is made up of experts in the fields of AI, computer science, privacy, law enforcement, civil liberties and public policy. The Board advises Axon around ethical issues relating to the development and deployment of AI-powered policing technologies and works to ensure these technologies ultimately serve the communities where they will be used.
Helping lead the Board is the Policing Project at New York University School of Law, a nonprofit organization that partners with communities and the police to promote public safety through transparency, equity and democratic engagement. The Policing Project works with the Board to facilitate its meetings and publish the Board’s conclusions and recommendations.
“At present, there are very real concerns about the accuracy of face recognition, and particularly about biases in the way it identifies people along racial, ethnic and gender lines,” said Policing Project Founder and Faculty Director Barry Friedman, the Jacob D. Fuchsberg Professor of Law at NYU, who is also a member of the Board. “Until we address and mitigate these challenges, we cannot risk incorporating face recognition technology into policing. The Ethics Board applauds Axon for acting consistently with this.”
The report released today was the Board’s first publication.
Whether face recognition on body-worn cameras ever can be ethically justifiable is an issue the Board has begun to discuss, and will take up again if and when the technology becomes more reliable and equitable.
In addition to Friedman, the Board includes Ali Farhadi from the University of Washington; Christy E. Lopez from Georgetown Law; Jeremy Gillula, technology and civil liberties leader; Jim Bueermann, formerly of the National Police Foundation; Kathleen M. O’Toole, formerly of the Seattle Police Department; Mecole Jordan from the United Congress of Community and Religious Organizations; Miles Brundage from OpenAI; Tracy Ann Kosa from Stanford Center on Philanthropy and Civic Society; Vera Bumpers from the National Organization of Black Law Enforcement Executives and Sheriff Walt McNeil of Leon County, Florida.
“We know that the bias evident in the policing of communities of color is unfortunately not missing from today’s artificial intelligence and facial recognition products,” said Mecole Jordan, from the United Congress of Community and Religious Organizations in Chicago. “The Board agreed that facial recognition technology has a disproportionately negative impact on women and people of color. Incorporating it into body cameras too soon could intentionally and unintentionally exacerbate tensions between some police departments and marginalized populations.”
“Policing works better when informed by thoughtful research,” said Jim Bueermann, founder of the Future Policing Institute, former president of the National Police Foundation, former police Chief and long-time advocate of evidence-based policing. “The research is telling us that face recognition isn’t in a stage of readiness to be used effectively on body cameras. Until face recognition can accurately help law enforcement officers identify individuals, the Board agrees that it should be kept off body cameras.”
The full report of the Axon AI & Policing Technology Ethics Board is available at policingproject.org/axon. For more information and interviews please contact firstname.lastname@example.org or (212) 992-6950.
The Policing Project at New York University School of Law is a nonprofit organization that partners with communities and their police to promote public safety through transparency, equity, and democratic engagement. For more information, please visit policingproject.org.