U.S. Rep. asks NIST to develop demographic standards for facial recognition tech
A member of the U.S. House of Congress has called on the National Institute of Science and Technology (NIST) to create a framework for the development and use of facial recognition technology in order to correct the potential for discriminatory results.
Rep. Emanuel Cleaver (D-Mo.) addressed a letter (PDF) to NIST Undersecretary of Commerce for Standards and Technology Dr. Walter G. Copan, and begins by acknowledging the transformative potential of facial recognition and other advanced technologies.
“I write today with optimism for the potential of technology to continue to improve society, but with the understanding that government must play its part in guiding the use of technology – including facial recognition technology – down optimal paths,” Cleaver explains.
Problems related to bias and widely varying degrees of accuracy have continued to be examined by U.S. government bodies, including in hearings held earlier this year by the House Oversight and Government Reform Committee’s Subcommittee on Information Technology.
With the possibility of Amazon’s Rekognition being deployed by ICE, presumably for use with diverse populations, Cleaver suggests that customers of the technology should be well informed with testing standards. The ACLU has requested information from ICE parent agency DHS about its plans for facial recognition use.
“To that end, I am requesting that NIST design and endorse industry standards and ethical best practices for independent testing of demographic based error rates in facial recognition technology,” Cleaver writes. “Some facial recognition technology vendors do not participate in Face Recognition Vendor Tests or in Face in Video Evaluations, and demographic based discrimination in algorithms is often unnoticed by developers unless effectively included in testing.”
Cleaver also requests information on NIST data sets used by facial recognition developers, and the steps NIST takes to ensure their appropriate degree of demographic diversity and reflection of operational conditions. In addition to protecting subjects from bias, the establishment of standards could also benefit government and private sector customers of the technology, Cleaver suggests.
Earlier this year, Cleaver wrote to the Department of Justice to request it investigate whether the use of facial recognition violates civil rights safeguards.