Facial recognition technology has been around since the 1960s, and it is the current focus of California Assembly Member Phil Ting’s proposed assembly bill – The Body Camera Accountability Act (AB 1215). AB 1215, if enacted, would prohibit law enforcement agencies from using facial recognition technology in connection with a body camera. While civil rights and liberties groups are championing the passage of AB 1215, California’s law enforcement lobby is urging California to allow the law enforcement community to work with tech groups to find a way to use facial recognition technology to protect California’s 40 million residents.
How Does Facial Recognition Technology Work?
Briefly, facial recognition software locates distinctive features on the face and the measurements of the facial features, such as the distance between eyes. These measurements are compiled to create an algorithm or biometric template of a person’s face. Facial recognition software then stores the template of the facial image in a database and compares the template to other stored images. Law enforcement agencies began to use the semi-automated technology by the 1980s, and one of the most well-known examples of law enforcement using facial recognition technology is the Lakewood Division of the Los Angeles County Sheriff’s Department. In 1988, the Lakewood Division used images of suspects captured on surveillance tapes and compared those images against its database of mug shots to find matches.
Now, the technology is fully automated and allows for real-time face recognition, which many Californians enjoy when using platforms such as Facebook as it allows them to quickly identify friends and family in photographs that are uploaded to the platform. In the public sector, the federal government has been using real-time face recognition since 2002 at the Statute of Liberty as all visitors pass face recognition cameras that compare their faces to a database of terrorist suspects. Additionally, the US Customs and Border Protection uses the technology at departure gates at airports in the US. The program, Biometric Exit, includes a face-matching system that is intended to “replace the need to manually check paper travel documents by providing an automated identity verification process…” per the CBP. This is just two of several examples of how the public sector is using the technology.
If The Technology Is Already Being Used By In Both The Public And Private Sector, Why Should Law Enforcement Agencies Be Prohibited From Using It?
With AB 1215, law enforcement agencies would be banned from installing and using facial recognition and biometric scanners in police body cameras. For example, the ban would prohibit body cameras from scanning an individual’s face and comparing the breakdown of the face to data in an internal police database (e.g. a mugshot database or the FBI’s database which currently has 640 million photos) to determine whether the individual should be detained. The stated concerns are (1) false identifications which could open innocent individuals to wrongful arrest. Axon, a police body camera manufacturer, has taken heed of this specific concern and announced in June 2019 that it will not use the technology until it can “accurately help law enforcement officers identify individuals.” An additional concern is (2) “subject[ing] law-abiding citizens to perpetual police line-ups, as their every movement is tracked without consent” which is a violation of one’s constitutional right to privacy per Ting.
Conversely, there have been significant advancements in facial recognition technology. The National Science and Technology Council Subcommittee on Biometrics and Identity Management stated that “every two years, the error rate in face recognition continues to drop by half.” Moreover, while in public, most citizens anticipate they will be observed to a degree, and there are few states that have laws in place that address whether an individual has a right to privacy over his/her biometric data while in public. Finally, law enforcement lobbies have pointed out that, facial recognition technology on body cameras will not extinguish the probable cause requirement. Thus, practically, police officers will not be wandering the streets simply scanning faces in public in order to make an arrest.
Given the tremendous strides that have been made with facial recognition technology, it would be worthwhile to at least explore whether the technology can be used with body cameras in a manner that affords greater security in California as opposed to a hardline ban based on concerns with issues that may be resolved.
 Facial recognition technology was initially “semi-automated,” meaning a system administrator had to locate key features in the photographs. Once the key features were manually identified, the system would calculate the distances from the key facial features and automatically compare the image to the reference data