WASHINGTON, DC – U.S. Senator Rob Portman (R-OH), Ranking Member of the Senate Homeland Security and Governmental Affairs Committee, introduced the Facial Accountability, Clarity, and Efficiency In Technology Act (FACE IT) to establish best practices and rules for the federal government’s use of facial recognition technology (FRT). FRT works by using algorithms to compare a person’s face to a photo of the person. It works by generating a mathematical representation of the person’s face using different attributes of a person’s facial structure and comparing that to the person’s face on file. By using AI, FRT has become exceptionally accurate. As a result, the federal government has deployed FRT in a number of law enforcement arenas including the Transportation Security Administration, Customs and Border Protection, and Federal Bureau of Investigation. However, there are few rules governing the use of FRT. The FACE IT Act aims to allow the federal government to continue the use of FRT systems, but with the addition of serious guardrails. As the FRT evolves, new rules may become necessary and so the legislation establishes an Advisory Commission on FRT to allow outside experts from industry and civil society to help the government ensure its continued safe use.
“Facial recognition technology can be used to help protect our communities, but I am concerned about the potential for abuse,” said Senator Portman. “I’m proud to introduce the FACE IT Act because, given the civil liberty implications of the federal government’s use of facial recognition technology, we must pass legislation to set rules for the use of this technology. We must make sure federal law enforcement and other agencies have the tools to do their jobs well, but it is vital that we set rules for those tools.”
NOTE: The FACE IT Act would establish best practices and rules such as: the establishment of accuracy requirements for the procurement of FRT, the requirement of human involvement in decision-making by FRT, and restrictions on federal agency’s access to FRT databases.