Stage 2 · Lesson 5 of 7
Probabilistic behaviour and what it means for security
The same prompt can produce different answers on different runs, because models sample from probabilities, and providers update models over time. For security this has three consequences. A defence that works on ten tries may fail on the hundredth, so results must be measured as rates, not single yes or no answers.
A test that passed last month may fail after a model update, so security tests have to run again on every change. And guards built from models share the same property: they are probabilistic too. This is why this roadmap keeps separating deterministic controls, such as code that checks permissions, which hold every time, from probabilistic ones, such as , which reduce risk.
Common trapreporting "I tried it and the defence held" after a handful of attempts.
A question to check yourself, an AI practice session and 1 cheat sheet entry come with the full roadmap. Unlock
Resources for this stage
Start with· follow one of these
- CoursePaidGrokking Prompt Engineering for Software Engineers
Design Gurus · Beginner · Subscription
How system prompts, context and structured output really work, the ground every LLM risk stands on
- CourseFreeHugging Face Learn
Hugging Face · Intermediate
Free courses on how language models, embeddings and agents work, at an engineer's level