AI + Metasploit = Terrifyingly Easy Hacking is here (demo)

簡介：

In this ethical lab demo, David Bombal and Kyle Winters connect Claude (LLM) to Metasploit through an MCP (Model Context Protocol) server to automate real attacks. Watch AI perform recon, generate a risk report, and execute VSFTPD backdoor, EternalBlue (SMBv1), and UnrealIRCD—dropping benign files on Linux and Windows with simple prompts. The demonstration is intended for educational use only on intentionally vulnerable VMs; do not attack systems you don’t own or lack permission to test. They show how MCP bridges an AI to real tools, how AI-driven scanning produces an auto‑generated security report, how prompted exploits are carried out, and discuss how this lowers barriers for red teams while outlining what blue teams should do. The video is structured with segments: 0:00 Coming up, 0:58 Disclaimer, 01:00 Introducing Metasploit MCP Server, 03:01 Metasploit MCP Demo 1, 05:12 Metasploit MCP Demo 2, 10:59 Metasploit MCP Demo 3, 16:18 Metasploit MCP Demo 4, 19:15 Metasploit MCP Demo 5, 21:45 How AI is changing cybersecurity, 23:07 Metasploit MCP Demo 5 continued, 26:51 Metasploit MCP server summary, 28:00 Conclusion.