AWS Certified AI Practitioner (AIF-C01) — Question 102

Which prompting attack directly exposes the configured behavior of a large language model (LLM)?

Answer options

• A. Prompted persona switches
• B. Exploiting friendliness and trust
• C. Ignoring the prompt template
• D. Extracting the prompt template

Correct answer: D

Explanation

The correct answer, D, refers to the act of retrieving the prompt template, which can directly reveal how the LLM is programmed to respond. Options A, B, and C either involve manipulating behavior without directly exposing the model's configuration or do not pertain to the attack on the model's behavior itself.