San Francisco Moves to Prohibit Government Use of Facial Recognition and Limit Surveillance Technology

Posted on February 23, 2019

According to a new piece by The Atlantic, San Francisco legislators are moving ahead with a bill that would prohibit city government agencies from using facial recognition technology. While the draft legislation, known as the “Stop Secret Surveillance Ordinance,” would not limit facial recognition owned and operated by private companies, a ban on the use of this technology would apply to all city government agencies. If passed, the bill would be the first and only total ban on government facial recognition technology in the country.

A key provision in the legislation also bars the city from obtaining facial recognition data, or data derived from facial recognition analysis, from third-parties. This language crucially closes a loophole by which the city could otherwise obtain facial recognition data from other jurisdictions, or from private sector partners who would still be free to use facial recognition commercially within city limits.

Moreover, the legislation goes beyond solely governing facial recognition surveillance in requiring approval by a board of supervisors before any other surveillance technology may be purchased and used by city government bodies. The ordinance would also obligate city agencies to file yearly reports on the effectiveness of surveillance technology in use, including all such technology in use by the city (even that which predates the law). This would extend to a broad range of software and devices potentially or actually in use by the city, from license plate readers to social media intelligence gathering programs to traditional CCTV cameras.

San Francisco’s pursuit of such an uncompromising protection against the extremely invasive monitoring posed by facial recognition surveillance could not be more timely. Many jurisdictions across the US are just now beginning to seriously debate mass adoption of the technology, often without real input solicited from, or transparency afforded to, the public.

As a CounterSpin interview (reprinted by FAIR) with ACLU Washington-Seattle’s Shankar Narayan on the subject noted, abuse of surveillance technology generally is not hypothetical, but historically well-documented. One need only look at the sprawling campaign, led chiefly by the FBI, to monitor and disrupt peaceful activists throughout the Cold War era. There is also evidence that license plate readers were deployed against Muslims after 9/11, demonstrating that surveillance abuse is not a relic of the 20th century.

As the CounterSpin piece points out, and as CCDBR has noted in past coverage, facial recognition, even when unbiased, can be easily harnessed to track the whereabouts of subjects caught on camera footage in real time by correlating the time and place of their image capture across the facial recognition network. Augmenting this real-time location tracking with the attribution of personal identities to surveillance subjects would undoubtedly prove an easy feat considering the existence of a nationwide database of faces encompassing the images of about half of US adults. The danger this poses is further compounded by the fact that facial recognition software is not a standalone product, as it can be integrated with public and private camera networks, enabled on body camera video footage, or even applied after-the-fact to video taken from practically any source.

Against this backdrop, San Francisco’s decision to tackle the issue at this early junction is extremely prudent, and will hopefully set the example for the rest of the nation.