NSF’s AI Agenda Focuses on Collaboration and Reducing Adoption Barriers

The National Science Foundation is working to address key challenges slowing artificial intelligence adoption as it positions itself as a collaborative and complementary partner in the Energy Department led Genesis Mission, according to agency leadership.

NSF has played a long standing role in advancing national technology priorities, including expanding access to research infrastructure, supporting open AI models for the scientific community, and investing in AI ready test beds. With the Genesis Mission led by the Department of Energy, NSF is refining how it can best support the broader effort.

Ellen Zegura, NSF’s acting assistant director for computer and information science and engineering, said the relationship between NSF and DOE continues to evolve. She noted that the agencies see opportunities to complement each other’s strengths while collaborating where missions align.

The Genesis Mission aims to establish a national AI platform and improve productivity tied to federal research and development investments. Zegura said NSF is focused on strengthening those goals while also identifying gaps not explicitly addressed by the executive order that launched the initiative.

One area of continued focus is workforce development. NSF released a roadmap outlining key skills and investments needed to build a strong AI ready STEM workforce and sought feedback from organizations and individuals across sectors. Lawmakers have also pushed to expand NSF’s role in this area through proposed legislation that would authorize AI related scholarships and fellowships.

Despite these ambitions, NSF faces the same challenges confronting many organizations adopting AI. Limited access to high performance chips and rising costs remain major obstacles. Demand for AI processors continues to grow rapidly, stretching global supply chains. Industry leaders have noted that existing cloud and GPU capacity is already fully utilized across multiple generations of hardware.

Cost pressures are another concern. Analysts estimate that AI tools often carry significant hidden expenses beyond initial purchase, including training, integration, and change management costs.

NSF is seeking a balanced approach that prioritizes access while remaining mindful of budget constraints. Zegura explained that while the agency may be priced out of the highest end AI capabilities, it can still play a critical role by supporting access at the next tier down, where many innovative and disruptive ideas emerge.

She emphasized that researchers need to be close enough to cutting edge resources to move ideas from early testing to larger scale applications. Without that proximity, the pathway to meaningful impact becomes much harder to sustain.

The agency has also undergone organizational changes over the past year, including workforce reductions and internal restructuring. These shifts have resulted in closer alignment between AI institute program officers and those overseeing advanced cyberinfrastructure. Zegura said this new structure could encourage experimentation with alternative AI architectures that may not have been possible before.

NSF’s AI institutes, first launched in 2019, continue to be viewed as a successful model. Over time, they have produced impactful research, education programs, and outreach efforts. While partnerships that support the institutes can be complex due to long timelines and rapid technological change, NSF is refining how those collaborations are structured.

Zegura said the agency is now focused on making more strategic investments and actively seeking input from research communities to guide future directions. As the national AI landscape continues to evolve, NSF aims to fill remaining gaps while supporting a broad and inclusive research ecosystem.

Submitted by Jason Gigax on Fri, 01/16/2026 - 16:29