Which risk of prompt engineering occurs when attackers insert malicious instructions into a model's context, such as data retrieved by RAG, so future prompts are influenced to produce incorrect or harmful answers?
Prompt hijacking
Prompt poisoning
Jailbreaking
Model drift
The described scenario is prompt poisoning. In a prompt-poisoning attack, adversaries inject harmful or misleading instructions into the data a foundation model later consumes (for example, documents indexed for RAG). When the model retrieves this tainted content, the malicious instructions become part of the prompt, leading the model to generate compromised outputs. Prompt hijacking instead refers to an attacker taking over an in-progress conversation, while jailbreaking tries to bypass a model's safety filters. Model drift is an unrelated performance change that happens gradually as data or requirements evolve, not from deliberate malicious instructions.
Bash is our AI bot, trained to help you pass your exam. AI Generated Content may display inaccurate information, always double-check anything important.
All IT & Cybersecurity Package plans include the following perks and exams .
Our pricing is simple. Full access to all certifications and exams in each package, for one price.
As many practice tests for as many topics as you want.
Use study mode non-stop, no limits.
Access to our AI assistant, Bash, trained to help you pass your exam.
Track your scores over time in study mode and report cards.
See how you improve over time, and where you need to focus.
Access our store with even bigger discounts than before.
Unlimited access to all performance questions and be prepared for the real thing.
All IT & Cybersecurity Package plans include unlimited access to the following study materials.
Create an account or sign in to access our study materials.