On The Record On Governing the Rules of AI
Dec 5, 2025
Join experts Laura DeNardis, a Georgetown professor specializing in technology ethics, Vinh Nguyen, a former NSA AI chief focusing on security, and Miriam Vogel, CEO of EqualAI and author, as they explore the intricate world of AI governance. They dive into the geopolitical challenges of AI, the pressing need for operational responsibility, and the delicate balance between innovation and accountability. From state-linked cyber threats to corporate responsibility, this discussion reveals what it takes to safely navigate the future of artificial intelligence.
AI Snips
Chapters
Books
Transcript
Episode notes
AI Is A Multi-Layered Stack
- AI is a broad, multi-layered field that includes models, chips, data, cloud, and security stacks.
- Treat governance as applying to an entire AI stack, not only the visible models.
Four Traits That Complicate Governance
- Frontier AI is hard to govern because it is probabilistic, opaque, feedback-driven, and can act autonomously.
- Those combined attributes create governance challenges we haven't previously faced together.
Operationalize Governance With C-Suite Ownership
- Monitor AI continuously and assign clear C-suite ownership to avoid fractured accountability.
- Build processes to find edge cases, test models, and ensure real-time stakeholder input.
