← Take me to the AgendaAI Agents and Automation

AI Composer - Safe Vibe Coding

Talk
Wednesday, April 1st 2026
14:55 - 15:15
Redford Stage

Description

LLMs generate code that appears correct yet omits critical security logic. In adversarial environments like DeFi, this is a structural failure, not a bug. This talk examines why GenAI cannot reason about adversaries, and why formal constraints are re