Artificial Intelligence-Ready Data Act
- Sponsor
- Sen. Budd, Ted [R-NC]
- Committees
- Commerce, Science, and Transportation Committee (primary)
- Last action
- Mar 16, 2026
Bottom line
This bill aims to standardize Federal open data to enhance its utility for artificial intelligence development across public and private sectors, with a specific focus on NOAA's forecasting data.
What it actually does
This bill directs the National Institute of Standards and Technology (NIST) to develop standards and guidelines for making open Government data assets "artificial intelligence-ready," focusing on aspects like data quality, interoperability, and privacy. It mandates that all Federal agencies adopt these standards and specifically requires the National Oceanic and Atmospheric Administration (NOAA) to ensure its data supports AI and machine learning for operational forecasting.
Proponents argue
Proponents argue that this legislation will unlock the immense potential of vast government datasets for AI innovation, significantly improve the efficiency and accuracy of government services such as weather forecasting, and foster economic growth by making data more accessible and usable for AI developers. They emphasize the critical need for consistent data quality and interoperability to maximize the benefits of artificial intelligence.
Opponents contend
Potential opponents might raise concerns regarding the substantial cost and administrative burden on Federal agencies to implement these new, comprehensive data standards, especially given the prevalence of legacy data systems. They could also highlight potential privacy and security implications of making more government data AI-ready, or the complexities involved in managing intellectual property when combining federal and proprietary data.
The bill is concise and clearly structured, allowing for a quick and straightforward understanding of its core objectives and mechanisms within the typical time allotted for legislative review.
Section 2, amending the National Institute of Standards and Technology Act by adding Section 37(a)
Establishment of AI-Ready Data Standards by NIST
This provision directs the National Institute of Standards and Technology (NIST), in consultation with other relevant federal agencies, to develop comprehensive standards and guidelines. These standards are intended to assist Federal agencies in making their publicly available data assets suitable for use with artificial intelligence, covering aspects such as data quality, metadata, and intellectual property management.
Supporters argue
Supporters argue that this provision is a crucial first step to harness the power of government data for AI, ensuring consistency and quality across agencies, which will accelerate AI development and innovation across various sectors.
Critics contend
Critics might contend that developing such broad and technically complex standards could be a lengthy and resource-intensive process, potentially delaying the actual availability of AI-ready data, or that a one-size-fits-all approach might not adequately address all agency-specific data needs.
Tradeoffs
The provision navigates the tension between the need for comprehensive, rigorous standards to ensure data quality and the desire for rapid implementation and sufficient flexibility to accommodate the diverse missions and data types of various federal agencies.
Section 2, adding Section 37(b)(3) to the NIST Act
Specific Requirements for AI-Ready Open Government Data
This section outlines specific criteria for open government data to be considered "artificial intelligence-ready." It mandates that such data must be easily downloadable (e.g., via publicly accessible websites or web-scraping), accurate, human-readable, in an open and machine-readable format accompanied by decoding tools, and secure while protecting individual privacy. These requirements are to be defined based on input from AI developers.
Section 3, adding Section 701 to the National Science and Technology Policy, Organization, and Priorities Act of 1976
Federal Agency Adoption and Adherence to AI Data Standards
This section mandates that once NIST develops and submits the AI-ready data standards, the President, acting through the Director of the Office of Science and Technology Policy (in consultation with OMB), must issue a requirement for all Federal agencies to adopt and adhere to these standards. Agencies are also required to ensure that any agency-specific adaptations remain interoperable and that major information technology and high-performance computing acquisitions explicitly account for these new data requirements.
Section 4
NOAA Data Readiness for AI/ML in Operational Forecasting
This section specifically directs the National Oceanic and Atmospheric Administration (NOAA) to ensure its data assets support the integration of artificial intelligence and machine learning into its operational forecasting activities. This includes critical data from model analyses, various observations (in-situ, conventional, satellite), and other environmental data. NOAA is also required to provide annual briefings to relevant Congressional committees on its progress for a period of five years.
NIST's standards and guidelines must recommend ways to manage intellectual property (IP) concerns when Federal data is combined with proprietary data.
Section 2, adding Section 37(b)(2)(A)(iii) to the NIST Act
Why it matters:This is a complex legal and policy issue that is often delegated to implementing agencies to figure out the specifics, rather than being detailed in primary legislation. Its placement here indicates it's a recognized challenge that needs to be addressed, but not the primary focus of the bill's high-level directives.
Case for: Proponents argue that clearly addressing intellectual property concerns is critical for fostering public-private partnerships in AI development, as private entities will be more willing to share or combine their proprietary data with government data if there are clear rules for IP protection and usage.
Case against: Critics might argue that defining and managing IP for combined datasets is inherently difficult and could lead to disputes, potentially creating barriers rather than facilitating data use, or that it could inadvertently favor certain private entities over others.
Estimated impact: Potentially significant impact on the scope and nature of public-private collaborations in AI, affecting data sharing agreements and the commercialization of AI models trained on mixed datasets.
Open Government data assets must be in an open and machine-readable format, *accompanied by publicly available software tools that can decode or process the data*.
Section 2, adding Section 37(b)(3)(E) to the NIST Act
Why it matters:This is a crucial technical implementation detail that ensures true machine-readability and usability, going beyond just specifying a data format. It's a practical necessity for AI developers but might not be highlighted in a high-level summary of the bill.
Case for: Proponents argue that merely providing data in a machine-readable format is insufficient if the tools to interpret that format are not readily available. This provision ensures practical accessibility for AI systems, significantly reducing the burden on developers.
Case against: Critics might point to the additional burden and cost on federal agencies to develop, maintain, and make publicly available software tools for every dataset, especially for specialized or evolving data formats.
Estimated impact: Enhances the practical usability of federal data for AI developers, potentially reducing friction and accelerating model training, but could incur additional development and maintenance costs for agencies.