CREATE AI Act
- Sponsor
- Sen. Young, Todd [R-IN]
- Committees
- Commerce, Science, and Transportation Committee (primary)
- Last action
- Apr 29, 2026
Bottom line
The CREATE AI Act establishes a federally coordinated national resource to democratize access to AI research infrastructure and foster AI talent development across the United States.
What it actually does
This bill establishes the National Artificial Intelligence Research Resource (NAIRR) within the National Science Foundation (NSF) to provide U.S. researchers, educators, and students with access to computational, data, software, and educational resources for AI research and skill development. It outlines the governance structure, including a Program Management Office, defines eligible users, mandates research security protocols, and allows for a fee schedule with a guaranteed free tier. The NAIRR aims to accelerate AI innovation and build a diverse AI workforce.
Proponents argue
Supporters argue that the NAIRR is crucial for maintaining U.S. leadership in AI by providing equitable access to essential resources, which are currently concentrated in large tech companies. They contend it will accelerate scientific discovery, foster innovation, and build a diverse AI workforce by enabling researchers from various institutions, including small businesses and minority-serving institutions, to participate in cutting-edge AI development.
Opponents contend
Critics might raise concerns about the potential for bureaucratic inefficiencies in managing such a complex resource, the cost implications for taxpayers, and the risk of the NAIRR not keeping pace with rapidly evolving private sector AI capabilities. Some may also question the effectiveness of a government-led initiative compared to market-driven innovation, or express concerns about data privacy and security within a centralized resource.
The bill is concise and can be read and understood within a reasonable timeframe, allowing for informed evaluation.
Section 2, amending the National Artificial Intelligence Initiative Act of 2020, adding Sections 5602(a) and (c).
Establishment and Governance of the National AI Research Resource (NAIRR)
This provision mandates the National Science Foundation (NSF) to establish the NAIRR within one year of the bill's enactment. It also requires the NSF to create a Program Management Office (PMO) within its Office of Advanced Cyberinfrastructure to oversee the NAIRR's budget, coordinate resources from various entities (government and non-government), and set policies and procedures.
Supporters argue
Supporters argue that centralizing the management of the NAIRR within the NSF, a leading science agency, ensures expert oversight and leverages existing infrastructure. The PMO structure is designed to provide necessary coordination and strategic direction for a complex, multi-stakeholder initiative.
Critics contend
Critics might contend that establishing a new office and resource within an existing agency could lead to bureaucratic bloat or mission creep, potentially diverting resources from other NSF priorities. Concerns could also be raised about the efficiency of a government-run entity in a fast-moving field like AI.
Tradeoffs
The tension lies between the desire for centralized, coordinated access to AI resources and the potential for bureaucratic overhead or slower adaptation compared to a purely private sector approach.
Section 2, amending the National Artificial Intelligence Initiative Act of 2020, adding Sections 5603 and 5604(a).
Scope of Resources and Access for U.S. Researchers
This provision defines the resources the NAIRR will coordinate and provision, including computational resources, data, software, models, testbeds, and educational materials. It also specifies who can access these resources, primarily U.S.-based researchers, educators, and students affiliated with institutions of higher education, non-profits, government agencies, federally funded R&D centers, and small businesses. It explicitly excludes individuals employed by foreign countries listed as adversaries.
Section 2, amending the National Artificial Intelligence Initiative Act of 2020, adding Sections 5604(b) and (c).
Security Protocols and Open Source Encouragement
This provision mandates that the NAIRR conform to National Security Presidential Memorandum-33 (NSPM-33) regarding research security and integrity. It requires the appointment of research security points of contact within the PMO and any delegated operational entities. Additionally, it establishes policies to encourage software developed for or with the NAIRR to be open-source.
Section 2, amending the National Artificial Intelligence Initiative Act of 2020, adding Sections 5604(d) and (e).
Funding Mechanisms and Oversight
This provision allows the PMO to establish a fee schedule for NAIRR access, which may vary by user type but must include a free tier and ensure the primary purpose is research support. Fees collected can be retained and used by the PMO without further appropriation. It also mandates the establishment of an Advanced Cyberinfrastructure Advisory Committee to assess and make recommendations on the NAIRR's activities, with specific membership requirements.
Requires conformance with National Security Presidential Memorandum-33 (NSPM-33) and its implementation guidance on research security and integrity.
Section 5604(b)(1)
Why it matters:Standard legislative practice to reference existing executive policy rather than re-stating it in full. It ensures consistency with broader national security directives.
Case for: Ensures that the NAIRR's operations are aligned with comprehensive national research security policies, protecting sensitive AI research from foreign interference and intellectual property theft.
Case against: While necessary, relying on external documents can make the bill's full implications less transparent to a casual reader, requiring additional research to understand the specific security burdens and restrictions.
Estimated impact: Significant, as NSPM-33 outlines extensive requirements for disclosure, foreign influence mitigation, and research integrity, impacting all aspects of NAIRR operations and user conduct.