Facial biometrics’ law enforcement potential debated around the world
As facial recognition becomes more commonly used, controversies and calls for limitations continue around the world.
Both sides of recent mass protests in Hong Kong have been using facial recognition to identify individuals on the other side, according to an interview with New York Times reporter Paul Mozer by PBS.
Protesters also began taking pictures of police officers, after they deployed without wearing visible badgers, to identify them using facial recognition. One protester, who Mozur says developed his own facial recognition tool for identifying police, was reportedly targeted by officers who attempted to unlock his phone by holding it in front of his face, but he had disabled the facial verification unlock feature. Mozur also notes that protesters have spray painted the lenses of CCTV cameras black in at least one instance.
New York police plan to continue using facial recognition, but only for investigations, a representative said during a panel discussion hosted by the NYU Center for Urban Science and Progress, and reported by the Brooklyn Eagle.
NYPD Detective Bureau Assistant Chief Jason Wilcox told the panel on “The Promises & Dangers of Facial Recognition,” organized by the Downtown Brooklyn Partnership, that people expect the force to use cutting-edge technology, and that it is “not scanning the people walking down the street, trying to identify wanted people.”
New York Civil Liberties Media Democracy Fund Technologist Fellow Jonathan Stribling-Uss argued that the police are not using the technology in as limited a way as suggested, according to its own documents, citing a recent report from Georgetown Law’s Center on Privacy and Technology that revealed the use of an image of actor Woody Harrelson to identify a criminal look-alike.
Wilcox emphasized that this unusual use was an outlier among 7,000 requests made to the force’s facial unit last year.
An editorial in the Indian Express by a pair of researchers from the Centre for Internet and Society sounds the alarm over India’s plans to launch an Automated Facial Recognition System (AFRS) capable of integration with the country’s passport and National Automated Fingerprint Identification System databases.
Even the Crime and Criminal Tracking Network and Systems (CCTNS) database that the AFRS will initially be linked to risks over-representing marginalized communities, they argue. The researchers say the database has colonial origins, and a connected network of CCTV cameras will present a large potential attack surface for hackers. They express concern about a lack of consent by people being analyzed, potential loss of privacy and other civil liberties, and doubts about the technology’s effectiveness.
“The experience of law enforcement’s use of FRT globally, and the unique challenges posed by the usage of live FRT demand closer scrutiny into how it can be regulated,” the researchers conclude. “One approach may be to use a technology-neutral regulatory framework that identifies gradations of harms. However, given the history of political surveillance by the Indian state, a complete prohibition on FRT may not be too far-fetched.”
Civil liberties advocates in Britain are also calling for public facial recognition to be banned, while police say it is a potentially highly valuable tool, according to a separate PBS report.
Hannah Couchman, lead spokesperson on facial recognition for the group Liberty, calls facial recognition “enormously dangerous to modern democracy.”
Digital Barriers CEO Zak Doffman, however, says the privacy lobby has skewed the statistics to create a misleading narrative in the media. He also says that law enforcement agencies are beginning to understand the importance of image quality, and that when used properly, the technology is highly effective.
Doffman agrees, however, with conservative MPs who have called for police trials to be paused while Parliament brings the legal landscape up to speed.
An opinion piece by Duane Blackburn for The Hill seeks to clarify several issues relating to facial recognition, such as the need to use different metrics for performance depending on how the system operates. He argues that it is one of the most powerful tools for combatting child sex trafficking, and that it is extremely accurate for some applications, while open-set identification is much more difficult.
Blackburn is a Science and Technology Policy Analyst at the non-profit MITRE Corporation, former White House Office of Science & Technology Policy Assistant Director, and developer of NIST’s Face Recognition Vendor Test in 2000.
He recommends the creation of a group of experts in face recognition, civil liberties, and privacy who are neither advocates for or against the technology’s use, to guide legislators, policymakers, and the press. He also says similar risks from other technologies must also be understood, and that privacy and civil liberties laws need to be updated.