This week, California became the latest state to block the use of face recognition technology (FRT) on police body-worn cameras amid concerns of surveillance and misidentification. The new law follows similar moves by New Hampshire and Oregon, as well as municipal bans in Somerville, Massachusetts; Oakland, California; and San Francisco, where Supervisor Aaron Peski described the technology as “fundamentally invasive.”
As public calls for legislation of FRT increase, we felt communities and governments could benefit from a broader understanding of what is currently being adopted or proposed. The Policing Project, with help from the Information Law Institute, conducted a review of enacted and proposed FRT regulation across the U.S. and Canada, culminating in our new resource that analyzes trends in legislation and provides a table of nearly 200 proposed and enacted laws.
Trends, gaps and what you should know about FRT regulation
Face recognition technology is increasingly being used by government agencies, particularly law enforcement. The problem is that the technology is not always accurate—in fact, it has significantly higher error rates for women and people of color. Further, law enforcement use of FRT is sometimes secretive, and some argue that, in the hands of the state, facial rec raises the specter of surveillance of political activities, protests, and engagement in other protected forms of expression and association. Those concerns have only been magnified by recent videos of Hong Kong protesters wearing masks and tearing down facial recognition cameras.
Meanwhile, though regulation of FRT is increasing, the rollout has been far from uniform. Many of the new and proposed laws focus on the use of the technology by law enforcement, but FRT is also being utilized by other state actors, including school districts and public housing authorities, as well as private businesses, from concert venues to department stores. For example, even in San Francisco where the government is now banned from deploying face recognition technology, landlords could still use it to control who gains entry to a building.
Our review found a wide variety of enacted and proposed legislation across the U.S. and Canada that, broadly speaking, fell into three categories: general regulations that ban, pause, or study FRT; operations-based regulations that control how FRT is deployed; and data-based regulations that restrict the images used to operate FRT. In our new resource, we explore the general characteristics of each of these categories, not to suggest that proper regulation should be confined to one category or another, but rather to provide any jurisdiction considering regulating FRT with a basis on which to begin studying existing measures.
The regulation of FRT is one of the most pressing technological issues facing governments today. Our hope is that with this information in hand, legislators and the broader public will be empowered to pursue regulatory options that are responsive to the needs of their communities and that sufficiently account for the potentially significant privacy and civil rights implications facial recognition presents.
Explore our new resource, The Growing World of Face Recognition Legislation.