AAISM: Advanced in AI Security Management

By Isaca · Question Mode

QuestionQ75

AI Technologies and Controls

An organization has deployed a natural language processing model to answer customer inquiries when personnel are unavailable. A security assessment performed before implementation found that attackers could obtain sensitive company data through a chat-interface injection attack. Which of the following is the BEST method to prevent this attack?

Explanation

Chat-interface injection attacks use untrusted user content to manipulate a model’s instructions or cause disclosure of protected data. Input validation and controlled templates help separate and constrain untrusted input, limiting attempts to override instructions or elicit sensitive information.

Community Discussion

No comments yet. Be the first to start the discussion!