Keynote – Trust But Verify The Evolving Risk Profile of AI Generated Code

This talk features Tom Howlett discussing the implications of using Large Language Models (LLMs) for code generation in production environments. He presents a framework to understand varying AI coding behaviors and explores the evolving risk profile associated with generated code. The session emphasizes the importance of governance and verification in light of increasingly powerful AI tools, providing vital insights for those responsible for implementing AI coding solutions at scale.