Why Top Security Researchers Are Thriving with Agentic AI

This article profiles a successful bug bounty researcher, Hacktus, describing how he leverages AI as a force-multiplier for reconnaissance while retaining human judgment for exploitation and business-logic flaws; it argues that AI expands attack surface and production of vulnerable code, highlights common weakness patterns around AI plumbing (e.g., prompt injection, authorization confusion), and offers advice for researchers to target new fields, automate processes, and focus on judgement-driven testing.