A closed-door, filmed roundtable of senior voices across AI, cybersecurity, government and data infrastructure, covering what it takes to move from a cyber-search world to a cyber-analytic one — and why getting there matters beyond the balance sheet.

The dominant conversation about AI and security is the wrong one. The acute risk is not AI-empowered attackers running at machine speed; it is the AI organizations are deploying internally: models, agents, copilots, and the data flowing in and out of them. Hash matches and signature lookups cannot tell you whether an agent is behaving the way it should. Studying behavior can — and that requires treating non-traditional security data as first-class, at a scale most current architectures cannot ingest or analyze without the cost curve becoming unsustainable.