DC416 — AI Fight Club: Talk + Workshop

Why we picked it

Everyone is shipping "AI-powered" security. Almost no one is honest about where it works and where it guesses. This month, two DC416 organizers run the same fight from opposite corners. The Thursday talk frames it. The Friday workshop puts it in your hands. THE TALK Amir Hosseinpour covers offense. A year of pointing AI at real production code, plus everything DC416 has tried as a community. Where it actually finds bugs, where it just burns tokens and lies to you about it, and the scaffolding that gets any frontier model behaving the same way regardless of vendor. Cost numbers. The kind that make you reconsider whether that last scan was worth it. And an honest answer to when you should put the LLM down and read the code yourself. KC Udonsi covers defense. Most blue-team tooling was built assuming a human reads every alert. AI breaks that assumption in ways that are not obvious until something goes wrong. KC built his own AI-native breach detector and is running it live, in the room