Accelerating R&D for Critical AI Assurance and Security Technologies

This is a linkpost for a policy memo published by the Federation of American Scientists outlining a strategic, coordinated policy approach to supporting R&D to address urgent assurance and security challenges relating to frontier AI systems.

The memo makes the following recommendations:

• OSTP and NSF should clearly define AI assurance and security R&D priorities by incorporating critical AI assurance and security areas into the 2025 update of the National AI R&D Strategic Plan, the forthcoming AI Action Plan, and guidance across all federal agencies.

• Federal agencies should establish an AI R&D consortium and deploy agile funding mechanisms for critical R&D. This consortium would bring together government, AI companies, and universities and leveraging innovative approaches like fast grants and prizes to accelerate R&D progress, with Congress supporting expedited funding for prioritized areas.

• NSF should establish an AI Frontier Science Fellowship to ensure a pipeline of technical AI talent that can contribute directly to R&D and support fast-grant program management. This would train early-career researchers in critical AI assurance and security while building administrative capacity for rapid grant distribution.

→ Click here to read the memo on the Federation of American Scientists’ website.