Root Cause

     

AI, Cyber and National Security

October 25th, 2025
Chris Rohlf

This week I had the pleasure of guest lecturing on Artificial Intelligence, Cyber and National Security at both Andrew Imbrie's course at Georgetown University and Ben Buchanan's course at Johns Hopkins SAIS in Washington DC.

A single lecture is not enough time to cover all aspects of where AI and cyber intersect in ways that impact the national security mission, but I tried to cover a number of important topics that are relevant to these courses. These topics included:

  • The defenders dilemma, how it creates an attacker favored asymmetry in cyber, and how AI has the potential to reverse this dynamic
  • How AI provides scale and automation in the vulnerability lifecycle
  • The threat model and technical approaches behind securing AI infrastructure, agents and systems
  • Why technical enforcement of semiconductor export controls requires a secure cryptographic foundation
  • AI 2035: A thought provoking scenario involving how AI may disrupt the security dilemma and create escalation risk between states

You can find the slides I presented here. As always the questions and discussion in person were more valuable than any slides I can prepare. I hope to refine this presentation based on that discussion and deliver it again in the future.