We’ve all been there: you press TAB, the AI assistant writes 20 lines of flawless-looking code, and you feel like a 10x engineer. But did you just invite SQL injection to the party?

AI coding assistants are revolutionizing software engineering, but they are also incredibly efficient at automating the OWASP Top 10. This talk explores the security blind spots of AI-generated code.

We’ll look at real-world examples of how LLMs autocomplete vulnerabilities, analyze the shifting responsibility of the modern developer, and discuss practical strategies to ensure your new silicon colleague doesn't accidentally tank your security posture.