Government Agents: AI Agents Meet Tough Regulations
summary
Los Alamos builds an agent that reads papers, generates hypotheses and writes simulation code, then executes those thermodynamic and hydrodynamic runs on Venado (≈2,500-node Grace Hopper system) to design an ICF capsule that optimizes yield — and the pipeline actually produces and simulates a viable slice-through design. Work includes bringing OpenAI models into classified networks, partnerships with UC, Nvidia and HPE, and chem/bio safety collaborations. Compliance hurdles: OMB memoranda M-25-21/M-25-22, FedRAMP, NIST SP 800-53 rev4, NIST AI RMF, DoD SRG/CNSSI 1253; requirements: explainability, isolation/IL5 deployability, SBOM and continuous monitoring.
AI-written summary · about
chapters (6)
- 0:00Introduction and background
- 1:14AI agents for scientific acceleration
- 3:20Government AI challenges and scale
- 6:09New AI governance requirements
- 9:43Future of AI governance in government
- 15:00Conclusion and national security impact
related talks
40mAgents in Action: Advancing Open Source AI Missions
35mThe Fight for the Future of Artificial Intelligence
01h 42mBuilding Responsible AI Agents in Decentralized Open-Source Environments
34mAI Agents: Transforming Autonomous Decision-Making
49mAI Agents: Collaboration in Real-Life Applications
34mThe Hard Truth About AI Agents in Production
14mReliable Agentic AI, the Open Way
07h 25mAI Agents Conference 2026 - Day 2