AAISM: Advanced in AI Security Management

AAISM: Advanced in AI Security Management practice questions, page 19 of 19, with answer review and explanations.

Question 91

Which approach should an organization prioritize to effectively verify the security of its AI models?

  1. Implementing an algorithm to automate the identification of vulnerabilities within the AI system
  2. Developing a testing strategy that includes AI-specific threat modeling and adversarial attack simulations
  3. Testing AI development team competencies in IT threat and vulnerability mitigation techniques
  4. Utilizing regular penetration testing methods similar to standard IT applications

Question 92

When deriving statistical information generated by AI systems, which of the following sources of risk is MOST important to address?

  1. Presence of hallucinations
  2. Incomplete outputs
  3. Lack of data normalization
  4. Systemic bias in data sets