Title
Calling on the Philadelphia Police Department, as well as state and federal law enforcement officials, to establish clear boundaries and issue transparent policies regarding the use of facial recognition technology to ensure that this technology does not lead to racial biases in policing practices and outcomes or infringe on individuals’ civil liberties.
Body
WHEREAS, Advances in artificial-intelligence (AI) technology have yielded new tools capable of scanning individuals’ faces, comparing them with databases of photos online in real-time, and providing accurate identification to law enforcement. The federal government and many cities and states have experimented with the use of these new law enforcement tools, raising concerns about their potential for civil liberties abuses and the entrenchment of problematic racial biases in law enforcement; and
WHEREAS, Philadelphia Police have confirmed reports of the Department’s trial run of a controversial facial recognition software, Clearview AI, which is capable of using surveillance cameras to compare an individual’s facial data to a database of billions of images the company mined from social media sites to provide law enforcement an accurate identification of anyone who passes in view of the camera. The Department has for years utilized another facial recognition software, JNET, which employs a more circumscribed database of pictures pulled from official sources such as drivers’ license bureaus and state and county prison databases; and
WHEREAS, Law enforcement has employed facial recognition software to advance a variety of pressing public safety concerns, including the ongoing battle against the opioid epidemic. Noting the important role that PPD’s JNET system played in securing the arrest of individuals involved in a major heroin and fentanyl distribution scheme, PPD Sgt. Andrew Callaghan claimed “this would not have happened without the magic of JNET”; and
WHEREAS, This automation of law enforcement tactics leaves open the concerning possibility that the software’s underlying framework would serve to exacerbate already existing biases in law enforcement practices. Several studies have shown concerning disparities in the accuracy of facial recognition software when it is used on individuals of color and women. Such disparities at the time of identification have the potential to lead to grave errors at the expense of these already disadvantaged groups; and
WHEREAS, Many experts have also raised broader civil rights concerns about the potential for AI facial recognition software to enable ubiquitous surveillance at the expense of individual liberties and public political expression. The ability to identify and track individuals in public, even where there is no reasonable suspicion of a crime, has the grave potential to create a chilling effect in public political participation at the expense of our democratic institutions; and
WHEREAS, A variety of proposals have been floated as a means of ensuring that facial recognition software is employed in a racially neutral manner and with due respect for individuals’ civil liberties. Some states have proposed regulations requiring third party testing of facial recognition technology in advance of deployment to ensure that the software does not generate biased law enforcement strategies or outcomes. Independent verification, both in advance and during the deployment of any facial recognition software, can serve to assure the public that law enforcement is striking the proper balance in its efforts to keep our communities safe; and
WHEREAS, The public’s interest in equipping law enforcement agencies with the most up-to-date crime fighting tools must be balanced against the grave privacy and civil rights concerns that these technologies present. Well-informed regulations that provide meaningful protections against these outcomes are necessary if our City, state, and federal law enforcement officials are to pursue an effective, modern public safety strategy in the 21st century; now, therefore, be it
RESOLVED, THAT THE COUNCIL OF THE CITY OF PHILADELPHIA, Hereby calls on the Philadelphia Police Department and state and federal law enforcement officials to establish clear boundaries and issue transparent policies regarding the use of facial recognition technology to ensure that this technology does not lead to racial biases in policing practices and outcomes or infringe on individuals’ civil liberties.
End