Secure Agentic Development for the Non-Expert Operator

Would you manually inspect hundreds of parts an hour in an assembly line? No, you'd build the inspection into the line. Just about anyone can now direct an AI agent to create code and ship software at a breathtaking speed, but not many can personally vouch that the result is safe and secure. This is the agentic development security assurance gap, and we'll discuss how to close the gap with the workflow itself through machine-enforced controls, adversarial AI reviews, and human checkpoints where it counts.

Panda is an AI/ML patent-holding security engineer with 7+ years securing systems at a Fortune 50 company.