The Biden administration on Tuesday unveiled a new artificial intelligence agenda intended to shape the way federal agencies and other institutions use the rapidly evolving technology.
President Biden’s “Blueprint for an AI Bill of Rights” is kick-starting a push from federal agencies making recommendations it wants others to follow across society, from schools to health care facilities.
The White House plan is focused on a handful of issues, mainly discrimination, privacy, efficiency and safety, explanation and notice, and providing human alternatives to automation.
Education Secretary Miguel Cardona said the blueprint released Tuesday is the foundation for guidance his department will distribute to schools to follow next year.
“While we embrace utilizing edtech to enhance learning, we recognize that it comes with additional responsibility and the need for us to change how we do business,” Mr. Cardona said at the White House. “We’re going to be releasing next year, early ‘23, a set of guidance and recommendations for schools, for 50 million students across the country.”
It is not just the Department of Education readying new directions; the Health and Human Services Department is doing likewise.
HHS Secretary Xavier Becerra said his department is analyzing artificially intelligent systems in hopes of producing a report later this year for people in the health care sector to learn whether they are making appropriately equitable decisions.
“We know that for the most part the folks who are left out of the health care system are low-income and people of color, and so if you use data that you have an abundance of, you probably are leaving out the most key elements and targeting everyone,” Mr. Becerra said at the White House. “At HHS, we’re doing work right now on equity essentially by design and we’re doing an analysis of what AI systems are out there and what they really mean to us.”
While health and education officials prepare the new directives, some parts of the Biden administration have already made changes to the government’s use of AI.
For example, the blueprint published Tuesday said the Transportation Security Administration is implementing a “gender-neutral algorithm” for its body-scanner operators to use at airport security checkpoints in response to complaints from transgender people.
“Body scanners, used by TSA at airport checkpoints, require the operator to select a “male” or “female” scanning setting based on the passenger’s sex, but the setting is chosen based on the operator’s perception of the passenger’s gender identity,” the blueprint said. “These scanners are more likely to flag transgender travelers as requiring extra screening done by a person. Transgender travelers have described degrading experiences associated with these extra screenings.”
The blueprint published Tuesday is neither new law nor an executive order, but a white paper from the White House’s Office of Science and Technology Policy.
The office believes more rules will likely be necessary, however, as the paper said the office expects “sector-specific guidance” to create rules for automated systems in various places such as “part of school building security or automated health diagnostic systems.”