The Cybersecurity Defenders Podcast

Agentic SecOps Workspace (ASW) office hours with LimaCharlie / Defender Fridays [#288]

Jan 30, 2026
Live office-hours demo of AI agents performing automated malware decompilation and analysis. AI-driven rule tuning at scale, including noisy detection investigation and autonomous false-positive fixes. Automated infrastructure tasks like tenant provisioning, data source setup, and GitSync configuration. Discussion of permission models, safety controls, and practical use cases for agentic SecOps.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
ANECDOTE

AI-Driven Headless Reverse Engineering

  • Maxime used Cloud Code to install Ghidra, parse docs, and build a CLI wrapper so the AI could run headless binary analysis.
  • He fed a WannaCry sample and the agent produced a decompilation-based report with IOCs and behavior analysis.
INSIGHT

Pattern Reasoning Beats Variable Names

  • The AI can reason about decompiled code even with opaque variable names because it recognizes code patterns and function relationships.
  • Maxime warns against trusting the AI for threat attribution without corroborating threat intelligence.
ADVICE

Corroborate Attribution With Threat Intel

  • Do not rely solely on AI for threat attribution; corroborate with external threat intelligence sources.
  • Integrate your MCP server or RAG-enabled threat intel so the agent can compare findings before attributing actors.
Get the Snipd Podcast app to discover more snips from this episode
Get the app