ISACA Certified AI Security Manager — Question 23

A large language model (LLM) has been manipulated to provide advice that serves an attacker's objectives. Which of the following attack types does this situation represent?

Answer options

• A. Data poisoning
• B. Evasion attack
• C. Privilege escalation
• D. Model inversion

Correct answer: A

Explanation

The correct answer is A, Data poisoning, as this attack involves injecting misleading data into the training set of a model to influence its outputs. The other options, such as Evasion attack, Privilege escalation, and Model inversion, pertain to different types of vulnerabilities and manipulations that do not specifically involve altering the training data to mislead the model.