CyberWire Daily

Logging off in Myawaddy.

Oct 29, 2025
Ben Seri, Co-founder and CTO of Zafran, delves into the rising threat of AI-native attacks, emphasizing how attackers are leveraging AI for vulnerability discovery and malware creation. He explains how organizations can harness generative AI for improved vulnerability management and remediation, suggesting a balanced approach with human oversight. Seri also cautions about the risks of in-house AI development and the potential for new exposures. Plus, he highlights the innovative use of agentic AI to simulate impacts and automate responses in security.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Attackers Move Fast With AI

  • Attackers rapidly adopt AI to scan, craft, and exploit vulnerabilities at scale.
  • Defenders must match that pace or risk falling behind as adversaries automate offenses.
INSIGHT

LLMs Turn Text Into Action

  • LLMs excel at parsing unstructured text and turning analysis into actionable plans.
  • That strength maps directly to vulnerability triage and remediation planning in security workflows.
ADVICE

Begin With Human-In-The-Loop

  • Start with human-in-the-loop controls when deploying agentic AI and define clear vendor guardrails.
  • Gradually automate only the actions that reach a proven accuracy threshold to reduce hallucination risks.
Get the Snipd Podcast app to discover more snips from this episode
Get the app