What is the purpose of the proposed National Frontier AI Access Allocation, and how should access be structured according to the chapter?

The proposed National Frontier AI Access Allocation is meant to give researchers at universities, national laboratories, and nonprofit institutions access to advanced AI systems at a scale sufficient for sustained experimentation, method development, and validation. Access should be broad-based, lightweight to obtain, and designed to support open-ended exploration as well as defined research projects, treating AI usage like a shared national research resource analogous to telescope time or supercomputing hours.

According to the chapter, the federal government should establish this allocation so that scientific researchers across universities, the government's own national laboratories, and nonprofit institutions can use frontier AI systems for sustained research work. The scale of access must be enough for ongoing experimentation, method development, and validation rather than one-off use. Structurally, access should be broad-based rather than narrowly restricted, lightweight to obtain, and intended to support both open-ended exploration and defined projects. The chapter explicitly compares this to other shared scientific resources such as telescope time or supercomputing hours, implying allocation should function as a common national research resource for the scientific community.