Lesson 1148 of 1596
Jailbreak Categories: Mapping the Adversarial Surface
Jailbreak attacks fall into recognizable families — role-play, encoding, persona, multi-turn pressure. A category map drives durable defense.
Creators · AI Foundations · ~7 min read
The premise
AI can map jailbreak categories and defensive postures, but your specific safety policy must define what counts as a successful attack.
What AI does well here
- Generate per-category jailbreak example sets for red-team use.
- Draft defensive-posture summaries by category.
What AI cannot do
- Define what content your platform considers harmful.
- Substitute for ongoing red-team practice.
Key terms in this lesson
Practice this safely
Use a small project example from your own work. The useful move is to compare the AI's draft against your goal, sources, and constraints before you trust it.
- 1Ask AI to explain jailbreak in plain language, then underline anything that sounds uncertain or too broad.
- 2Give it one detail from "Jailbreak Categories: Mapping the Adversarial Surface" and ask for two possible next steps plus one reason each step might be wrong.
- 3Check role-play attack against a trusted source, teacher, adult, expert, or original document before you use it.
End-of-lesson quiz
Check what stuck
10 questions · Score saves to your progress.
Tutor
Curious about “Jailbreak Categories: Mapping the Adversarial Surface”?
Ask anything about this lesson. I’ll answer using just what you’re reading — short, friendly, grounded.
Progress saved locally in this browser. Sign in to sync across devices.
Related lessons
Keep going
Creators · 32 min
Jailbreak Mechanisms and Defenses: How Adversaries Bypass AI Safety
Jailbreaks exploit prompt-format, role, and capability gaps; understand the mechanism categories to evaluate vendor defenses critically.
Creators · 11 min
Attention deep dive: queries, keys, values, and why it works
Understand attention as a content-addressable lookup over a sequence — and where the analogy breaks.
Creators · 11 min
Tokenization economics: why your bill depends on the tokenizer
Tokenization decisions ripple into cost, latency, and capability — for languages, code, and rare strings.