On Nov. 7, representatives from federal agencies, academia and the private sector gathered at the U.S. National Science Foundation to kick off a collaborative process to design a pilot program for a National Artificial Intelligence Research Resource (NAIRR). NAIRR is a concept for a shared national research infrastructure that will connect U.S. researchers to responsible and trustworthy AI resources, as well as the needed computational, data, software, training and educational resources to fuel AI research and discovery.
"We are excited about the expanded community of innovation that is emerging from NAIRR, and the pilot convening has been a tremendous success."
- NSF Director Sethuraman Panchanathan
On Oct. 30, President Biden signed Executive Order 14110 directing NSF to launch the NAIRR pilot program by bringing together federal and private sector contributions to expedite federal government support for responsible AI research and the U.S. AI research community, particularly researchers in academia and small businesses. In this way, NAIRR seeks to democratize access to AI innovation and support critical work advancing trustworthy AI.
The event fostered a merging of insights from prospective providers of the resources and expertise that will be made available through the pilot program. Participants discussed the dual aims of the pilot to advance AI research and democratize access, as well as how to quickly leverage existing resources and best practices. This gathering is the first in a series of planned workshops and meetings that will engage the broader community in the design of the pilot, from bringing in user perspectives to setting guidelines and principles for the responsible use of the platform and the safety of the research conducted through it.
The planning for the NAIRR pilot program is informed by an implementation plan that was submitted to the president and Congress in January 2023, developed through a two-year consultative process by a task force co-chaired by NSF and the White House Office of Science and Technology Policy. Congress created the task force as a federal advisory committee to address the growing accessibility gap between the most well-resourced institutions, laboratories and companies and the broader AI research and education communities in terms of computational data and software resources. The NAIRR pilot is intended to bridge this gap, providing broad and equitable access in a manner that protects privacy, civil rights and civil liberties. The program is also intended to enable the research community to pursue responsible and trustworthy innovation with safeguards that mitigate risks and potential harms, in alignment with the federal government's commitment to advancing and upholding the principles of safety, security and trust in AI development.
"We are excited about the expanded community of innovation that is emerging from NAIRR, and the pilot convening has been a tremendous success," said NSF Director Sethuraman Panchanathan. "NAIRR builds on a legacy of successful public-private partnerships that have driven American leadership in research and innovation. We are proud to continue to welcome new partners to the NAIRR community in the months to come."
Implementing the pilot is just one responsibility charged to NSF in the recent Executive Order. Recognizing NSF's leading role in supporting AI innovation and cultivation of AI talent across the country, the Executive Order directs NSF to: expand its network of National AI Research Institutes, currently a $500 million investment to advance AI research across a range of application areas; establish an NSF Regional Innovation Engine to build an AI-focused innovation center outside of existing technology hubs; and prioritize resources to support AI-related education and workforce development through existing programs. In addition, NSF is directed to advance the development of privacy-enhancing technologies (PETs), including through the creation of a research coordination network dedicated to advancing the development, deployment and scaling of PETs. NSF will also develop testbeds and testing environments to support the development of secure and trustworthy AI technologies, as well as the design, development and deployment of PETs.
