Skip to content
Reference

YouTube Channels to Follow

YouTube Channels to Follow

Subscribe to these channels for ongoing AI security content. Organized by focus area — check back regularly as this field moves fast.

AI Security Specialists

  • 📺 Embrace The Red · 9.3K subs Johann Rehberger (former Microsoft MSRC). The leading prompt injection researcher — responsible for disclosures to Microsoft, Google, OpenAI. Regular attack demos against Claude Code, Copilot, Gemini. Recent: "Agent Commander: Promptware C2", "Breaking Claude Opus 4.7 Memory". ᴛᴏᴘ ᴘɪᴄᴋ — essential viewing for Phase 2.

  • 📺 SANS Institute · 68.5K subs AI Cybersecurity Summit talks, DFIR training previews, threat briefings. Long-form conference talks (30-60 min). Covers governance, IR, and defensive engineering across Phases 3-4.

  • 📺 Prabh Nair · 173K subs GRC, pentesting, and AI security masterclasses. Deep-dive format (1-2h sessions). Covers ISO 42001, CISO mindset, security architecture. Strong Phase 4 governance content.

Offensive Security & Hacking

  • 📺 John Hammond · 2.13M subs Malware analysis, supply chain attacks, AI-powered phishing. Weekly uploads covering real-world incidents. Recent: "npm axios supply chain attack", "ChatGPT For The Dark Web". Phases 2, 3.

  • 📺 David Bombal · 3.05M subs Interviews with top hackers + AI security demos. Recent: "Vibe Hacking: How AI Is Helping Hackers" (812K views), "Google Big Sleep: End of Human Hackers?" Strong general security + AI intersection.

  • 📺 LiveOverflow · 938K subs Deep technical security research and CTF walkthroughs. Pwn2Own documentary series. Less AI-specific but builds exploit development fundamentals needed for Phase 2.

  • 📺 NahamSec · 200K subs Bug bounty methodology, web hacking, recon automation. Practical offensive security that translates to AI app pentesting in Week 13.

  • 📺 PwnFunction · 238K subs Beautifully animated web security explainers (XSS, SSRF, prototype pollution). Only 22 videos but 8.2M total views — each one is premium quality. Builds the web vuln foundation for LLM output exploitation (Week 13).

AI/ML Education (Security-Relevant)

  • 📺 IBM Technology · 1.67M subs Polished explainers on AI security, agents, OWASP, MITRE ATLAS. Publishes 3-4x/week. Key videos: "OWASP Top 10 for LLMs" (168K), "Architect Secure AI Agents" (51K), "MITRE ATLAS" (24K). Covers all 4 phases. Short format (8-15 min) — good for commute watching.

  • 📺 Microsoft Developer · 676K subs AI Red Teaming 101 course, PyRIT tutorials, agent security patterns. Official Microsoft AI Red Team content. Key for Week 14.

Conference Channels

  • 📺 DEF CON Conference · 399K subs 4,600+ talks. AI Village sessions cover embeddings exploitation, LLM attacks, model security. Search "AI" for relevant talks.

  • 📺 Black Hat · 257K subs 3,000+ talks. Practical LLM security, enterprise AI red teaming, real-world vulnerability disclosures. Search "LLM" or "AI".

Shorts & Quick Content

Most AI security channels focus on long-form content (talks, tutorials, demos). For bite-sized AI security content:

  • 📺 IBM Technology — occasional 5-8 min explainers that work as quick primers (Agentic Security Runtime, Tool Calling, A2A Protocol)
  • 📺 David Bombal — produces some sub-10-min demos alongside longer interviews (e.g. "AI malware demo" at 9:39)
  • 📺 Prabh Nair — occasional 4-min quick hits alongside deep dives

Note: True YouTube Shorts (<60s vertical video) are rare in the AI security space — the field is too technical for that format. The channels above that do shorter content use standard horizontal video in the 5-15 min range.