Professor Jennifer Laurin testified at a virtual hearing focusing on the use of Facial Recognition Technology by law enforcement.
Professor Jennifer Laurin testified during a virtual hearing before the U.S. House of Representatives Subcommittee on Crime, Terrorism, and Homeland Security on July 13, 2021. The hearing, titled “Facial Recognition Technology: Examining Its Use by Law Enforcement,” aimed to generate not only an understanding of facial recognition technology (FRT) but also to proactively engage with it and seek legislative fixes to the problems it presents.
Chairwoman Sheila Jackson Lee opened the hearing by describing both the promise and peril of FRT. She asked whether the technology is sufficiently accurate to warrant widespread use in law enforcement, or if large-scale adoption could, in fact, inject further inequity into the criminal justice system.
Prof. Laurin’s testimony addressed an important aspect of this complex topic that intersects with her research and teaching on forensic science and criminal investigation and adjudication.
Watch the hearing (Laurin begins at the 53:50 mark)
In her testimony, Prof. Laurin presented three bottom-line points. First that, “…although FRT ‘matches’ have not been held to be admissible in evidence in criminal cases, law enforcement use of FRT to identify suspects forms the basis for arrests, criminal convictions, and periods of incarceration in the federal system.”
Prof. Laurin testified that the use of FRT does lead to deprivations of liberty, because evidence of FRT matches are regularly relied upon by courts during the stages of criminal cases not subject to the rules of admissibility. She further stated, “even where FRT is used by law enforcement only to generate an initial lead, that early FRT match has the potential – particularly given the stickiness of cognitive biases – to send investigators down an erroneous path that, even if ultimately corrected, has enormous consequence for a wrongly accused individual.”
Prof. Laurin’s second point argued, “while FRT has the capacity to be highly accurate under ideal conditions, there is greatest cause for concern about accuracy when used in criminal investigations.”
She described the specific concerns about FRT accuracy when used in criminal investigations asserting, “Studies of FRT that find low error rates do so only when using cooperative, front-facing, staged images; declines in accuracy are consistently seen when using side-view images, low-resolution images, or other images not captured under controlled circumstances – precisely the types of images that are often used in criminal investigations in which the perpetrator’s identity is not yet known.”
Lastly, she elucidated her third point, “that the criminal legal system is not well-designed to screen out mistaken or unreliable fruits of facial recognition technology.” Defendants are extremely limited in their ability to litigate law enforcement use of FRT, and many never even learn that an early FRT match procedure was performed. The fact that many of the systems are proprietary makes vetting of the technology even more difficult.
Additional hearing testimony was given by NYU School of Law professor and Faculty Director of their Policing Project, Barry Friedman, Gretta Goodwin, Director of Homeland Security and Justice in the U.S. Government Accountability Office, and others.
Prof. Laurin is the Wright C. Morrow Professor of Law at The University of Texas. Her scholarship focuses on the role of constitutional litigation in regulating police, the shared roles of courts, police, and lawyers in regulating forensic science, and oversight of indigent defense. She currently serves as Reporter to the American Bar Associations’ Criminal Justice Standards Task Force.