BPC Applauds the Roadmap for Artificial Intelligence Policy in the U.S. Senate

The Bipartisan Policy Center applauds Senate Majority Leader Chuck Schumer and Sens. Mike Rounds, Martin Heinrich, and Todd Young for their thoughtful bipartisan leadership in crafting their “Roadmap for Artificial Intelligence Policy in the U.S. Senate.” BPC recognizes the profound implications—both positive and potentially negative—this swiftly advancing technology holds for our society and economy. The working group’s Roadmap carefully balanced these potential outcomes, ensuring the United States remains the clear leader in AI development and deployment.

AI in Government

We agree that the federal government should focus on getting its data ready for AI use and have advocated for this initiative to spur innovation and access to data. Government initiatives should focus on creating an all-of-government “AI-ready data” infrastructure, ensuring that data is accessible and usable for AI applications. Moreover, directing research priorities towards responsible innovation in fundamental and applied sciences—such as biotechnology, advanced computing, robotics, and materials science—along with trustworthy AI topics like transparency, explainability, privacy, interoperability, and security will foster a robust and ethical AI ecosystem.

The importance of funding the National Institute of Standards and Technology (NIST) efforts for this roadmap to be effective cannot be overstated. We are heartened by the funding language that acknowledges the necessary resources to realize the promise of responsible AI. As appropriations are considered, it is crucial to recognize that funding for AI should be separate from general funding for IT in the federal government. This distinction ensures that the work to modernize federal IT from a cybersecurity perspective remains a fundamental and separate focus from AI modernization.

Recruiting, retraining, and retaining an AI-ready workforce are priorities we are pleased to see included in this roadmap. Utilizing the US Digital Service, the Presidential Innovation Fellows, and the Presidential Management Fellows is a great start. Focusing on bringing talent to the federal workforce should continue to be a priority. We encourage other mechanisms to be considered, such as an AmeriCorps focus on AI-Literacy, a program to detail AI proficient employees to agencies standing up AI programs, and an expansion of the statistical agencies’ workforce who are already equipped to endeavor on data-heavy research projects.

Finally, we appreciate the roadmap’s callout for a comprehensive data privacy law. Data privacy has been an outstanding concern for legislators and constituents for years. AI is only going to scale up those concerns and considerations. Unless we address the policy issues in the technology stack that enables AI, we will always have unanswered questions and liability in future system deployment. A federal data privacy law must be established to alleviate many privacy questions as AI systems are deployed.

Integration of AI in Health Care

We agree that prioritizing patient needs is crucial when incorporating AI into health care. Legislation supporting the integration of AI in health care must include guardrails to protect patients and ensure the use of accurate data. We support the roadmap’s call for strong data management, recognizing the importance of accurate and diverse data as well as patient privacy protection in AI development. While there are established regulatory frameworks currently being used for certain AI applications, there are no specific guidelines for the emerging field of generative AI. The roadmap correctly notes that the Department of Health and Human Services (including the Food and Drug Administration and the Office of the National Coordinator for Health Information Technology) needs to have the right tools to develop a predictable regulatory framework for AI-enabled devices.

As lawmakers develop legislation, it’s important to examine the use of AI in health care’s administrative and backend functions to clarify existing oversight and identify necessary improvements.

BPC’s Health and Technology Programs support activities to ensure companies can safely address AI-related risks to children’s mental health in product design— especially for social media platforms— to further research their impacts.

AI in Efficient Tax Administration

We are encouraged by the working group’s emphasis on updating IT infrastructure with modern data science and AI technologies to improve the delivery of government services. BPC’s latest research explores how this could significantly enhance the delivery of tax and benefit programs by improving customer service and reducing the administrative burdens associated with program verification and eligibility. Policymakers should create guidelines on the utilization of AI for tax administration specifically—whether to assist and answer taxpayer questions or to manage and analyze large amounts of data, such as screening for improper payments—with clear policies on data protection. Coupled with proper training for staff, guidelines will help ensure that agencies secure and use data appropriately as they integrate the information into their technological infrastructure and modernization plans.

AI in Workforce Training and Retention

We are pleased that the impact of AI on the workforce is a top priority for the working group and their emphasis on the importance of training, retraining, and upskilling workers. We further encourage the working group and other policymakers to prioritize lifelong learning and consider a broad range of approaches that can be tailored to the specific needs and circumstances of workers. These approaches can include high-quality internships, apprenticeship programs, and innovative credentialing methods. Furthermore, we encourage the working group and policymakers to explore ways to promote AI literacy, ensure proper AI governance in the workplace, and evaluate the adequacy of social safety net programs in the context of AI-driven job disruption. Relatedly, BPC supports bipartisan legislative efforts to update our immigration system to attract and retain top talent in AI and other critical technology fields.

Department of Energy (DOE) AI

We applaud the working group for identifying methods to address the rising energy needs of AI. The development of appropriate metrics for and greater transparency on data center energy use will help ensure we meet our nation’s energy needs without curtailing AI innovation and growth. Beyond advancing solutions to power AI computation, DOE has deep expertise in AI R&D as well as building and operating supercomputers, recently announcing that the national labs are home to the top two fastest supercomputers in the world. Additionally, DOE recently announced the Frontiers in Artificial Intelligence for Science, Security and Technology (FASST) initiative will harness DOE’s AI expertise to advance the agency’s energy, science, and national security missions. Given DOE’s long history and breadth of expertise on AI, we encourage Congress and the working group to think more strategically about how DOE and the national labs can advance a national AI strategy. A successful U.S. strategy for driving AI innovation must properly leverage and build upon existing federal programs and resources, and DOE is central to the success of this mission.

BPC is enthusiastic that the working group considered the views of so many individuals, organizations, and businesses. Such a broad, multistakeholder approach safeguards against shortsighted or knee-jerk policies that could harm Americans or stifle the development of cutting-edge technologies. We remain hopeful that the working group’s bipartisan leadership will help spur innovation and drive economic growth.