The US National Science Foundation has unveiled a $100 million initiative to establish state and regional AI infrastructure hubs designed to support scientific research and address geographic disparities in compute access. The programme, formally titled NSF State and Regional Artificial Intelligence Infrastructure Hubs, will initially support up to ten consortia with one award made per state or region.
Each hub will function as what the NSF describes as a flexible consortium of state and local governments, research institutions, philanthropies and private-sector participants. The federal funding will cover consortium coordination, workforce development and faculty training, while the consortia themselves will be responsible for funding new or expanded compute infrastructure. The structure allows hubs to pool resources across states and institutions of different sizes.
The initiative responds to recommendations in a July 2026 report published by White House Office of Science and Technology policy director Michael Kratsios. That report identified obstacles facing US researchers and outlined recommendations centred on four main pillars, though the NSF documentation does not detail those pillars further.
Brian Stone, performing the duties of the NSF director, said the programme supports a national mission to advance AI for science. "Artificial intelligence is transforming how we conduct research, accelerate scientific discovery and address complex challenges across disciplines," Stone said. "The State and Regional AI Infrastructure Hubs program supports a national mission to advance AI for science by expanding access to critical AI resources, strengthening state and regional ecosystems, and developing the workforce capabilities needed to harness these technologies."
Stone added that the effort aims to "empower researchers, students, and communities to advance AI-enabled science and drive the next generation of scientific breakthroughs." The NSF framed the programme as a mechanism to work with regional industrial partners to support both scientific and local workforce needs.
The foundation is actively seeking partners to support the hubs. A number of private-sector and philanthropic organisations have already signalled their intent to participate, according to the NSF. Those include chipmakers Nvidia, AMD and Intel, as well as Dell Technologies, Hangar and the Secunda Innovation Fund.
The consortium model places capital-raising responsibility on the state and regional participants rather than the federal government. NSF funding covers operational and human-capital development costs but leaves infrastructure investment to the local partnerships. That structure could open co-investment lanes for family offices and other private capital sources willing to underwrite compute assets alongside public and academic institutions.
The programme announcement arrives as AI infrastructure investment has become a focal point for both institutional and private allocators. The NSF initiative adds a regional-consortium layer to a buildout that has been dominated by hyperscale cloud providers and large technology companies. Whether state-level hubs can achieve the scale and utilisation rates required to compete with centralised infrastructure remains an open question.
The timeline for hub selection and the specific criteria for awards were not detailed in the NSF documentation. The foundation's emphasis on geographic equity and workforce development suggests that allocation decisions will balance technical capability with regional economic and educational impact.