AI is changing security fast ..... but is it replacing pentesters, or just giving them a terrifying new co-pilot? In this episode, we get into the real future of offensive security: what AI can automate, what it absolutely can’t (hello, business logic flaws), and why prompt injection + hallucinated packages are the kind of problems that keep AppSec people emotionally unwell. 03:46 — Will AI replace pen testers? 06:05 — What AI can automate now (tier-1 recon, discovery, noise) 08:16 — The human edge: business logic flaws (IDOR, logic bugs) 20:55 — How attackers target AI systems (prompt injection gets real) 23:55 — Hallucinated packages + “AI bias” (trusting it too much) 28:00 — AI agents + why you should only test them in a sandbox #CyberSecurity #Pentesting #AI #AppSec #LLMSecurity #PromptInjection #InfoSec