Jake built an autonomous AI scam agent: it texts its targets, chats back on its own, even rings them in a cloned voice and follows up by email. No human in the loop at any point

Then he let it loose: find out what happened when the AI agent went after real people, and why the results surprised even Jake

Then Jake became the agent: He stepped off the screen and played the AI's persona in person, walking into a real IT company with multiple AI tools working alongside me live, from identifying staff to breaking into the network

What it means for you: where AI agents makes scammers dangerous, where people still hold the line and what has to change before criminals attack at scale