Before cybersecurity was an industry, it was a scene — scrappy, rule-bending, and largely undocumented. In this fireside chat, journalist Joseph Menn shares decades of stories from reporting on that world, tracing how it shaped the practitioners defending modern systems. Menn pushes back on the idea that AI-accelerated bug hunting solves the security problem on its own. Someone still has to triage what gets found.

The throughline: critical thinking, not just faster tools, is what keeps defenders ahead.

With the explosion of generative AI, the fundamentals of cybersecurity don't change, but the cost of ignoring them goes up. Sanger argues that AI takes what is already possible in cyber and automates it, speeds it up, and makes it harder to trace, turning familiar risks into more urgent ones. From there, the conversation moves to why nuclear-era arms-control models don’t map neatly onto code and how automation is reshaping attribution in the new cold wars.

Securing critical infrastructure comes down to visibility, accountability, and judgment, even as AI speeds up detection and response. From tracking connected equipment to knowing who to call during an incident, the panelists agree that AI amplifies existing threats rather than replacing the need for human decision-making.

This panel on securing physical industries is moderated by Mary Writz, CPO at Bitwarden, and features Sean Habing (Director of Information Security, Pollard Banknote) and Anthony Green (CTO, FoxTech Cyber).