# Why AI Guardrails Don't Work: Cybersecurity Mental Models With Sounil Yu (Knostic AI)

Page: https://stenobird.com/podcast/tools-for-tech-leaders-7024721/why-ai-guardrails-don-t-work-cybersecurity-mental-models-with-sounil-yu-knostic-ai
Text version: https://stenobird.com/podcast/tools-for-tech-leaders-7024721/why-ai-guardrails-don-t-work-cybersecurity-mental-models-with-sounil-yu-knostic-ai.md
Podcast: [Tools for Tech Leaders](https://stenobird.com/podcast/tools-for-tech-leaders-7024721)
Published: 2025-06-12T11:18:00+00:00
Episode link: https://946a962b-4b7a-4ee9-a6ed-53f386ec1df5.libsyn.com/why-ai-guardrails-dont-work-cybersecurity-mental-models-with-sounil-yu-knostic-ai
Audio file: https://traffic.libsyn.com/secure/946a962b-4b7a-4ee9-a6ed-53f386ec1df5/TTL_-_Sounil_Yu_Podcast_audio.mp3?dest-id=4698167
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/tools-for-tech-leaders-7024721/episodes/why-ai-guardrails-don-t-work-cybersecurity-mental-models-with-sounil-yu-knostic-ai
Duration seconds: 2008

## Resource

Every AI guardrail gets broken, but that's not the real problem. Cybersecurity framework creator Sounil Yu reveals why traditional security approaches fail in the knowledge age and how enterprise AI deployments create dangerous "oversharing" vulnerabilities that most organizations don't even recognize. Learn the DIKW pyramid framework for understanding AI risks, discover why your security team's "allergy patterns" are sabotaging productivity, and understand the prisoner's dilemma forcing every organization to accelerate AI adoption despite known dangers. Essential insights for executives navigating the gap between AI opportunity and enterprise security reality.

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/tools-for-tech-leaders-7024721/episodes/why-ai-guardrails-don-t-work-cybersecurity-mental-models-with-sounil-yu-knostic-ai/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/tools-for-tech-leaders-7024721/why-ai-guardrails-don-t-work-cybersecurity-mental-models-with-sounil-yu-knostic-ai.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.