AAISM: Advanced in AI Security Management practice questions, page 17 of 19, with answer review and explanations.
Question 81
A PRIMARY objective of responsibly providing AI services is to:
ensure the confidentiality and integrity of data processed by AI models.
build trust for decisions and predictions made by AI models.
enable AI models to operate autonomously.
improve the ability of AI models to learn from new data.
Question 82
When integrating AI for innovation, which of the following can BEST help an organization manage security risk?
Evaluating compliance requirements
Re-evaluating the risk appetite
Adopting a phased approach
Seeking third-party advice
Question 83
A model producing contradictory outputs based on highly similar inputs MOST likely indicates the presence of:
evasion attacks.
poisoning attacks.
membership inference.
model exfiltration.
Question 84
An attacker crafts inputs to a large language model (LLM) to exploit output integrity controls. Which of the following types of attacks is this an example of?
Evasion
Jailbreaking
Remote code execution
Prompt injection
Question 85
Which of the following MOST effectively minimizes the attack surface when securing AI agent components during their development and deployment?
Schedule periodic manual code reviews.
Implement compartmentalization with least privilege enforcement.
Consolidate event logs for correlation and centralized analysis.
Deploy pre-trained models directly into production.