{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/6402f69f6a774eeab29a83db67aabcd1\" frameborder=\"0\" width=\"1920\" height=\"1440\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1440,"width":1920,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1440,"thumbnail_width":1920,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/6402f69f6a774eeab29a83db67aabcd1-82af0002ba7dc16b.gif","duration":438.074,"title":"AI Polygraph Autonomous Incident Response Agent","description":"This Loom presents AI-Pol, an AI Polygraph concept for an Autonomous Incident Response Agent for the Fine Evil Hackathon 2026. The system uses 13 models inspired by 30 animal defense and attack behaviors, with a Dreamcatcher-based theme for isolating “bad dreams” or cyber nightmares. In a Python simulation using Ollama with Llama, modules detect attacks such as missing authentication headers, brute force after three failed logins, and suspicious persistence, then generate an incident report with a score of 8 out of 10 and recommended SOC-TEM actions including isolation and logging to an immutable file. The example incident involves lateral movement and brute force, mapping tactics to T11-T12, privilege escalation T10-21 remote services, and exploit-related T11-90, with multiple isolations such as IP blocking and account lockdown."}