A statement issued by California’s American Union of Civil Liberties (ACLU) on Tuesday revealed that pictures of each state legislator of California were run through the newly developed program of facial-recognition. This program compares facial images with images in a database comprising mugshots of 25,000 criminals. According to the organization, the program made major mistakes by flagging more than 20 legislators as criminal suspects.
The results of the experiment were announced by the ACLU to press the passage of the bill framed with the intent of banning the technology’s use in body cameras of the police department. Member of the Assembly, Phil Ting stressed in the statement that the experiment was enough proof of the under-developed state of the software of facial recognition for use in prime time. Ting’s face was one of the 26 images that were flagged as mugshots. He reiterated that the software was far from being suitable of being used by the law enforcement cameras.
Systems dealing in facial-recognition have garnered popularity in the recent past. Currently, they are in use at schools, airports, concerts and homes. This technology is even being used to assist bartenders to identify the person next in the queue of getting a drink. The system functions by recognizing the faces of people from photos and videos. Then, it compares the facial features of the individuals to the ones stored in the database.
However, the ACLU’s concern lies with the inaccuracy and bias inherent to the technology, particularly with colored people and women and people of color. In collaboration with ACLU, Ting is the co-sponsor of AB 1215, which is termed as the Accountability Act for Body Camera. This bill seeks to ban the usage of systems involving facial recognition or any surveillance system using biometric data in body cameras worn by the police. At present, no city in California is using this technology for the body cameras of their police department.