The initiative emphasizes leveraging existing data rather than collecting new data to accelerate breakthroughs and strengthen U.S. leadership in AI-enabled science.

The plan reflects a shift in science funding toward risk-taking, infrastructure-based AI-enabled discovery, aligning with White House and OSTP policy discussions.

NSF plans to invest up to $100 million, with individual hub awards from $2 million to $5 million and planning grants up to $200,000.

The effort fits within a national strategy to unlock data value for AI, supporting structured datasets, public research platforms, and a future STEM workforce.

The National Science Foundation launches the State and Regional Artificial Intelligence Infrastructure Hubs program, a $100 million initiative to expand access to compute, data, and AI resources for researchers, students, and educators to enable AI-enabled scientific discovery.

The program aims to enable new AI-driven insights by building data pipelines, feature extraction methods, and metadata generation to make datasets usable for automated AI analysis.

Budget and policy context highlight NSF funding pressures and a push to reform grant procurement and encourage risk-taking in science.

The hubs will connect research, workforce development, and regional growth by aligning infrastructure with local employer needs and regional industries, fostering collaboration among universities, community colleges, government, and private industry.

The program integrates with national NSF initiatives and existing consortia to broaden workforce development and resource sharing across the AI ecosystem.

Projects may use existing NSF resources and platforms, including the NSF Integrated Data Systems and Services and the National AI Research Resource, plus platforms like DOE’s Genesis Mission-developed American Science and Security Platform.

The initiative builds on a prior NSF program to unlock dataset value for AI-enabled discovery by enhancing accessibility and interoperability of existing datasets.

Initial funding comes from a $100 million grant, with cross-state collaboration and shared resources to scale compute beyond what any single stakeholder could achieve alone.