

AI Threat Landscape: Model Poisoning and Prompt Injection
Episode notes
When you type a password, the computer knows it's data. But when you talk to an AI, your instructions and your data are the exact same thing—just tokens in a stream. That single flaw is the root of every AI attack. In this video, Sec Guy explains the math behind Universal Adversarial Triggers, reveals how Indirect Prompt Injection can turn a resume into a weapon, and shows why Token Smuggling allows malware to slip right past your firewall.
New Security+ SYO-801 Study Guide Available today: secguy.org/security-study-guide
[Exam Ready Route - FREE]
Pass your certification for $0.
✅ Training Videos & Practice Tests
✅ Sec Guy Mobile Lab (On-the-go training powered by AI voice)
✅ Discord Access (Study sessions & Industry networking)
👉 Start Here: ...