Preserving security in a world with powerful AI Considerations for the future Defense Architecture

TL;DR

The paper argues that the current U.S. defense program is ill-suited to preserve security amid rapidly advancing AI, and that security will require both fortifying legacy systems and creating new defensive architectures. It analyzes structural shifts in science and technology, including accelerated progress, democratization of threat capabilities, and potential rogue AI threats, to motivate a rethinking of deterrence and defense planning. To address these challenges, it calls for a large-scale threat-exploration effort, tighter public-private AI partnerships, and norms around extreme democratization, complemented by near-term actions for national laboratories, notably establishing an AI Factory for Defense Science and transforming laboratory workflows. The work aims to spur agile, cross-sector coordination and institutional reform that can provide early warning and adaptive defenses in an era dominated by powerful AI.

Abstract

Advances in AI threaten to invalidate assumptions underpinning today's defense architecture. We argue that the current U.S. defense program of record, designed in an era before capable machine intelligence, cannot by itself preserve national security against rapidly emerging AI enabled threats. Instead, shoring up legacy systems must be coupled with entirely new elements of a defense architecture. We outline immediate steps to adapt the Department of Energy National Nuclear Security Administration National Laboratories to ensure agility and resilience in an era of powerful AI.