Absent enforceability, public agencies and leadership plan on developing similar AI guidance and protocols to ensure accountable usage and design.
Leadership from various federal agencies discussed their individual plans to help enforce the blueprint for the first national artificial intelligence framework, following its formal release today. The framework contributes to the Biden administration’s goal of promoting responsible technological innovation in both the public and private sectors.
Echoing the AI framework’s key tenants, agencies like the Department of Education and Department of Health and Human Services discussed the importance of equity and anti-discrimination in AI algorithms and systems.
Education Secretary Miguel Cardona said that his agency’s stance is to embrace emerging technology as an asset to learning, but emphasized the need for “guardrails” and protections for students.
“We have to be cautious not to go into another system where we're not protecting our students protecting our families, especially those who might be subject to bias or stereotyping through technology,” he said.
Cardona clarified that protecting student data is paramount to implementing new technologies in education technology. He explained that the Department of Education will be releasing new guidance early next year that coincides with the new AI framework, to better educate families and teachers about how student data will be utilized in new education tech systems.
Department of Health and Human Services Secretary Xavier Becerra said his agency will also be releasing a report at the end of 2022 documenting how people of color tend to be harmed by algorithmic biases in the health care space. Given the pervasive nature of discrimination in algorithm-powered AIs, Becerra anticipated evaluating design transparency as a priority for his team.
“My sense is that we're going to be looking most for transparency, to try to make sure if you're going to use something we can…we understand how you developed it and how you're planning to implement it,” he said.
Becerra notably said that he would like to implement a rule to formally prohibit algorithm discrimination in health technologies “if the courts don't get in our way.”
“If we're able to move forward, we should be able to have a better look at how IT …is making use of AI, and we should have a good bible out there for folks to understand what's happened in the past and [with] discriminated-against communities,” Becerra said. “And we should also have a rule that sets the rules of the road moving forward on how to use health IT.”
These comments follow the Biden administration’s emphasis on the need for a whole-of-government approach to making the AI framework a formal standard for best practices in the artificial intelligence and machine learning space, especially given that the framework itself is not legally enforceable.
“These commitments are truly only a down payment,” Alondra Nelson, the deputy director for the White House Office of Science and Technology Policy said. “They are the start, the beginning of transformative change and the way government designs, uses and regulates automated systems. All of us have a role to play to bring tech development into the public sphere.”
NEXT STORY: Vote Now for the Future of Robotics