CCT 367: Threat Modeling and the AI Agent That Breached Hugging Face (CISSP Domain 1.10)

Episode 367

0:00-42:52

Send us Fan Mail A rogue AI agent didn’t “hack the future” so much as exploit the oldest security problems in the book: weak boundaries, over-trusted inputs, exposed endpoints, and credentials lying around. We walk through the Hugging Face intrusion story like a CISO briefing a board, step by step, translating a fast-moving AI-red-team narrative into the kind of clear threat modeling logic you need for ISC2 CISSP Domain 1.10 and for real risk decisions. From there, we shift into training mode and get concrete about threat modeling concepts and methodologies. We define threat modeling the way the exam cares about it: proactive, iterative, and designed to produce security requirements and prioritised countermeasures that feed your SDLC and risk register. We break down the three starting perspectives (asset-centric, attack-centric, and system-centric), then anchor STRIDE to data flow diagrams and trust boundaries so you can identify what security property is actually being violated, not just recite acronyms. We also cover the difference between identifying and ranking threats so you don’t fall into the classic traps: DREAD ranks threats you already found, while PASTA starts with business objectives and risk appetite and is built for risk-centric outputs leaders can fund.

Publisher

CISSP Cyber Training Podcast - CISSP Training Program

Published

August 24, 2026

Duration

43 min

Format

Podcast

Category

Lifestyle · Courses

Source

RSS feed

Host

Shon Gerber, vCISO, CISSP, Cybersecurity Consultant and Entrepreneur

CC

CISSP Cyber Training Podcast - CISSP Training Program

@cisspcybertrainingpodcastcis

#threatmodeling #breachedhuggingface #cisspdomain #courses

The Nile