RunAs Radio Hacking using AI with Erica Burgess
22 snips
Feb 18, 2026 Erica Burgess, an AI cybersecurity architect and former red team software engineer known for bug bounties and conference talks. She explores how LLMs spark unconventional exploit ideas and speed up coding and research. They discuss multi-agent orchestration, how AI lowers the bar for attackers, and implications for defenses and ethical prompting.
AI Snips
Chapters
Transcript
Episode notes
LLMs Accelerate Creative Exploitation
- Large language models can make creative leaps that speed up exploit development and generate untested but useful ideas.
- Erica Burgess found an LLM finished most of a remote code execution proof-of-concept that took her days to partially build.
Let AI Handle Tedious Recon
- Offload tedious research and enumeration to AI so humans can focus on creative exploit chaining.
- Use LLMs to generate obscure commands, tweak payloads, and iterate quickly on potential attacks.
Multi-Agent Hacking Orchestration
- Multi-agent orchestration yields diverse, parallel exploit strategies where specialists try different attack types.
- Erica composes manager and specialist agents to connect results and discover novel chains.