A federal task force has proposed a shared infrastructure for artificial intelligence research, aimed at giving more U.S. scientists access to the resources needed to work at the field’s frontiers. Its plan would bring computing power, data, software, testing tools and expert support together through an integrated portal.
Addressing a gap in AI research access
The National AI Research Resource (NAIRR) task force argues that advances in AI often depend on access to large amounts of computing power and data. Today, those resources are more readily available to well-resourced organizations, leaving other researchers with fewer ways to conduct ambitious work.
The report warns that this divide could narrow the range of people and organizations shaping AI research. If access remains concentrated, the research ecosystem may miss ideas and perspectives that could come from a broader group of scientists.
The proposed infrastructure is meant to widen participation and support research that addresses societal challenges. Its stated aim is to make AI research and development more accessible across the United States while advancing work that benefits people broadly.
A shared network rather than new data centers
Instead of beginning by building its own data centers, the proposed organization would assemble resources from partners. The report says a new organization should initially work with existing providers, which could include private companies or National Labs.
Researchers would be able to reach a federated mix of resources through one integrated portal. That mix could include computing and data resources, testbeds, algorithms, software, testing tools, networks, services and user support.
This approach would connect assets that are already available while avoiding the cost and scaling challenges of immediately establishing separate data centers. It would also make coordination central to the proposal: the organization would need to bring different resources and support services together in a way that researchers can use.
Governance, safeguards and the proposed budget
The task force recommends an independent research organization governed by the appropriate agencies and departments. It also says privacy, civil rights and civil liberties should be addressed from the outset through technical controls, policies and governance mechanisms.
Funding would require congressional action. The proposal calls for approximately $750 million every two years over a six-year period, totaling $2.25 billion to build out the resource. After that, the organization would need an estimated $60 million to $70 million per year for ongoing operations.
Those figures do not include related grants or similar programs. The report says those would likely be handled through the National Science Foundation or other existing programs, leaving the proposed organization focused on shared research infrastructure and support.
What happens next
The proposal sets out a direction and an implementation plan, but its practical details depend on money and other resources being assigned. The task force’s report provides more information about how the system could be run; the eventual arrangement would still depend on decisions about funding and partners.
The task force was established in 2020 and was headed by the White House Office of Science and Technology Policy. Its work included smaller reports and a blueprint for an AI bill of rights before this final report.
For researchers, the plan’s central promise is access: a way to draw on computing, data and expertise that might otherwise remain out of reach. Whether that promise becomes a working national resource depends on funding, coordination among providers and the safeguards built into its governance.