WASHINGTON, DC – Today, U.S. Senators Rob Portman (R-OH), Ranking Member of the Senate Homeland Security and Governmental Affairs Committee, and Martin Heinrich (D-NM), co-chairs of the Senate AI Caucus, along with U.S. Representatives Anthony Gonzalez (R-OH) and Anna Eshoo (D-CA), sent a letter to the Office of Science and Technology Policy (OSTP) and the National Science Foundation (NSF) clarifying congressional intent with respect to the National Artificial Intelligence Research Resource (NAIRR).
In 2020, the members introduced the bipartisan, bicameral NAIRR Task Force Act, which became law in the FY2021 National Defense Authorization Act. The legislation establishes a task force to develop a detailed roadmap for the development of a national AI resource for AI research and convenes a group of technical experts across academia, government, and industry to develop a detailed roadmap for how the United States can build, deploy, govern, and sustain a national research cloud and associated research resources. This letter reminds the administration of the congressional intent behind the NAIRR—to democratize access to AI resources and testing to make AI systems more safe and ethical—and offers steps the administration can take to better realize that intent.
“We introduced legislation to create the NAIRR Task Force because we see the incredible possibilities that the NAIRR offers for researchers across the United States. Smart individuals with good ideas should not need to work at a handful of large technology companies to have access to the computer power or other resources needed to research and deploy AI-based technologies. Our explicit aim for the NAIRR has been to democratize access to those resources, thereby diversifying the ranks of those working on cutting edge AI research, and making an investment in the next wave of American competitiveness,” wrote the members. “The United States will only reap the benefits of AI if it is deployed at scale. Yet, deployment at scale is only possible—and desirable—if Americans of all walks of life trust that AI systems are working in their interest. We are pleased by your shared commitment to improving that trust, and look forward to working with you to make the NAIRR a success.”
“In a recent survey conducted by Morning Consult, we found one area where Americans of all political affiliations agree: the federal government should ensure that their community has access to the resources needed to develop AI they can trust,” said Austin Carson, Founder & President of SeedAI. “Thanks to the leadership of Senators Portman and Heinrich, and Representatives Eshoo and Gonzalez, we are on the way to a National AI Research Resource (NAIRR). The NAIRR combined with an AI Bill of Rights, can be the perfect mechanism to begin putting what Americans want into practice.”
The text of the letter can be found below and here.
Dear Director Panchanathan and Director Lander,
As the bipartisan authors of the National Artificial Intelligence Research Resource (NAIRR) Task Force Act, we applaud your swift and faithful efforts to launch the statutorily required NAIRR Task Force and help chart a historic course for U.S. competitiveness. We also are pleased to see your efforts to design an Artificial Intelligence (AI) Bill of Rights to protect civil rights in our high-tech age, and believe there are unique synergies between the work of the Task Force and the development of the AI Bill of Rights. In reiterating the congressional intent undergirding the NAIRR Task Force, we encourage you to expand your ongoing efforts related to developing and deploying safe and ethical AI, and urge you to use the NAIRR Task Force as a valuable tool in those efforts.
We introduced legislation to create the NAIRR Task Force, because we see the incredible possibilities that the NAIRR offers for researchers across the United States. Smart individuals with good ideas should not need to work at a handful of companies on the West Coast to have access to the compute or other resources needed to pursue their ideas. Our explicit aim for the NAIRR has been to democratize access to those resources thereby diversifying—in a multitude of ways—the ranks of those working on cutting edge AI research, and making an investment in the next wave of American competitiveness. Understandably such a bold and unprecedented undertaking requires tapping into expertise across sectors, which is why the NAIRR Task Force is composed of representatives from government, academia, and industry.
While optimistic about the value of the NAIRR, we are concerned about the downsides posed by AI, especially with respect to civil rights and liberties. This is why we explicitly included an assessment of the privacy and civil rights and civil liberties requirements associated with the NAIRR and its research in the legislation, and hope that assessment by the Task Force draws from your work to articulate rules for the ethical use of AI with the AI Bill of Rights.
Additionally, the NAIRR should be used as a means to test and improve AI along the lines of the requirements associated with an AI Bill of Rights. As a federated, heterogeneous system-of-systems, the NAIRR should include a number of testbeds suited for conducting evaluations and research with implications for ethical AI across myriad disciplines and implementations. In this way the NAIRR can be used to operationalize aims of the AI Bill of Rights, while imbuing the AI Bill of Rights with the NAIRR’s core tenet of leveling the playing field for American’s access to technology.
To further this symbiosis between the NAIRR and the AI Bill of Rights, we believe it is vital to staff the NAIRR Task Force. Without staff, especially those with expertise on how AI resources can be used to support the development of ethical and safe AI, we are worried that it will be difficult to accomplish the exciting aims of the NAIRR, especially in the statutorily required amount of time. On the AI Bill of Rights side of the ledger, we urge you to add the right to participate in making and testing AI technology, including access to resources like the NAIRR. By taking these two steps, we believe you can better integrate the complementary aims of the NAIRR and AI Bill of Rights to the benefit of all Americans.
The United States will only reap the benefits of AI if it is deployed at scale. Yet, deployment at scale is only possible—and desirable—if Americans of all walks of life trust that AI systems are working in their interest. We are pleased by your shared commitment to improving that trust, and look forward to working with you to make the NAIRR a success.