Governments should:
- Set standards to require AI labs to promote safety
- Require licensing and reporting for very large training runs
- Security: support labs on information security, physical security, and vetting staff
- Develop compute governance
- Strengthen hardware export controls
- Establish software export controls
- Fund AI safety research
- Require and coordinate incident reporting
- Clarify liability of AI labs
General sources:
- Frontier AI Regulation (Anderljung et al. 2023)
- “Governmental levers” in “Advanced AI governance” (Maas 2023)
- AI’s Gatekeepers Aren’t Prepared for What’s Coming (Scharre 2023)