This is not a replacement for the official study guide. Please read the official study guide first — use this exam prep only to reinforce or practice your understanding.
Free AAISM exam practice questions with answers and explanations, organized by domain and part.
Domain 2: AI Risk Management
Q231 Which of the following types of testing can MOST effectively mitigate prompt hacking?
- A. Load
- B. Input
- C. Regression
- D. Adversarial
Answer: D
Prompt hacking manipulates large language models by injecting adversarial instructions into inputs to bypass or override safeguards. The AAISM framework identifies adversarial testing as the most effective way to simulate such manipulative attempts, expose vulnerabilities, and improve the resilience of controls. Load testing evaluates performance, input testing checks format validation, and regression testing validates functionality after changes. None of these directly address the manipulation of natural language inputs. Adversarial testing is therefore the correct approach to mitigate prompt hacking risks.
Domain 2: AI Risk Management
Q232 An attack has occurred on an AI system that has been in use for two years. Which of the following would BEST mitigate the impact of the attack?
- A. Monitoring AI systems for suspicious activities
- B. Updating deployed training data with new adversarial data
- C. Replacing the AI model with a new model that hides confidence levels
- D. Implementing strict access controls to the model’s architecture
Answer: B
When an AI system experiences an attack after being in production for an extended period, the most effective mitigation strategy is to update the deployed training data with new adversarial data. This process strengthens the model’s resilience by retraining it to recognize and resist attack vectors that were previously unknown or unaccounted for. According to the AI Security Management™ (AAISM) framework, risk mitigation for AI systems must address model robustness through adversarial retraining, data quality improvement, and model lifecycle hardening rather than relying solely on reactive measures. Why Option B is Correct: Incorporating adversarial examples into the training set enhances the system’s ability to correctly classify and withstand malicious inputs. This approach directly mitigates the vulnerability exploited in the attack and supports a proactive, continuous risk management cycle. Why Other Options Are Incorrect: Option A: Monitoring helps detect suspicious activity but does not resolve the underlying vulnerability. Option C: Concealing confidence scores may reduce model transparency but does not address the attack mechanism or its root cause. Option D: Implementing access controls protects the model’s architecture but does not improve model robustness against input manipulation attacks. Exact Extract from Official AAISM Study Guide: “AI risk management requires continuous improvement following incidents. After an adversarial or data poisoning event, the preferred risk treatment involves retraining the model using adversarial data and updated datasets to enhance robustness. This ensures the AI model adapts to evolving threat landscapes rather than merely restricting access or obscuring outputs.”
Domain 3: AI Technologies and Controls
Q233 Personal data used to train AI systems can BEST be protected by:
- A. Erasing personal data after training
- B. Ensuring the quality of personal data
- C. Anonymizing personal data
- D. Hashing personal data
Answer: C
AAISM guidance on privacy-preserving AI highlights anonymization as the most effective means of protecting personal data used in training. By irreversibly removing or masking identifiable attributes, anonymization ensures that training data cannot be linked back to individuals, thereby meeting key privacy obligations under laws such as GDPR. Erasing data after training may limit exposure but does not protect it during the training process. Ensuring data quality improves accuracy but does not mitigate privacy risk. Hashing protects data integrity but does not guarantee anonymity, as hashes can sometimes be reversed or correlated. Therefore, anonymization is the recommended control for protecting personal data in AI training.
Domain 2: AI Risk Management
Q234 An organization is updating its vendor arrangements to facilitate the safe adoption of AI technologies. Which of the following would be the PRIMARY challenge in delivering this initiative?
- A. Failure to adequately assess AI risk
- B. Inability to sufficiently identify shadow AI within the organization
- C. Unwillingness of large AI companies to accept updated terms
- D. Insufficient legal team experience with AI
Answer: C
In the AAISM™ guidance, vendor management for AI adoption highlights that large AI providers often resist contractual changes, particularly when customers seek to impose stricter security, transparency, or ethical obligations. The official study materials emphasize that while organizations must evaluate AI risk and build internal expertise, the primary challenge lies in negotiating acceptable contractual terms with dominant AI vendors who may not be willing to adjust their standardized agreements. This resistance limits the ability of organizations to enforce oversight, bias controls, and compliance requirements contractually.
Domain 1: AI Governance and Program Management
Q235 A board is told staff are already using public AI tools although no AI policy exists. What should the governing body do FIRST in response?
- A. Wait for an incident to justify funding AI governance
- B. Block all AI tools until regulators publish final rules
- C. Define an AI strategy that sets clear boundaries for AI use
- D. Delegate AI oversight entirely to the IT department
Answer: C
Staff use available AI tools whether or not a policy exists, so the governing body should set a strategy with clear boundaries first. Blocking everything or waiting for an incident leaves the risk unmanaged. The manual states governing bodies should define a strategy that sets clear boundaries for use.
Domain 1: AI Governance and Program Management
Q236 A loan-approval model begins rejecting qualified applicants after months of stable output. Which governance concern does this MOST directly illustrate?
- A. Excessive transparency exposing proprietary model internals
- B. Inadequate interface design hiding results from reviewers
- C. A shortage of GPU capacity limiting inference throughput
- D. Model drift degrading output quality and reliability
Answer: D
Models can drift or hallucinate over time, altering output quality, which is why governance must continue beyond initial compliance. Capacity or interface issues would not change decision quality for stable inputs. Ongoing monitoring and updating are part of AI governance.
Domain 1: AI Governance and Program Management
Q237 An executive asks why the AI governance program must go beyond legal compliance. Which response is MOST accurate?
- A. Ethical standards are the responsibility of AI developers rather than the wider organization
- B. Legal compliance is optional wherever no specific AI statute has yet been enacted
- C. Compliance with current law is sufficient because AI outputs then remain unbiased over time
- D. Governance must also sustain ethical standards and protect against reputational and legal damage
Answer: D
Current governance trends move from pure legal compliance toward AI social responsibility that protects against broader damage. Compliance does not guarantee fairness or ethical operation over time. Treating ethics as optional ignores reputational and harm risk.
Domain 1: AI Governance and Program Management
Q238 A credit team instructs an AI system to "defer payment until after the holidays." The system applies the instruction incorrectly. Which AI governance consideration does this BEST represent?
- A. A weakness in network segmentation between AI and production systems
- B. A lack of physical access control over the data center hardware
- C. The need to increase model training speed on larger datasets
- D. The need for explicit, precise instructions and understood system rules
Answer: D
An instruction clear to a human operator may be too imprecise for an AI system to execute correctly, so objectives and rules must be understood. The other options relate to unrelated control areas. This is the manual's example of transparency and explainability considerations.
Domain 1: AI Governance and Program Management
Q239 Which outcome is MOST likely when governance of AI is limited to the speed of existing human compliance controls?
- A. A growing gap between automated learning speed and control effectiveness
- B. A reduced need for ongoing workforce and skills planning across the business
- C. Automatic regulatory approval because manual controls are documented in detail
- D. Immediate elimination of model bias because humans review every output
Answer: A
The manual lists the growing disparity between the speed of change of automated learning systems and human compliance controls as a key consideration. Controls that move slowly cannot keep pace with adaptive systems. No other option follows from slow controls.
Domain 1: AI Governance and Program Management
Q240 An organization wants to confirm it is ready to adopt AI. Which action is the BEST first step?
- A. Announce AI adoption plans publicly to signal innovation to customers
- B. Assess current readiness to use AI with adequate governance
- C. Train a proprietary foundation model on internal data to gain advantage
- D. Purchase an enterprise AI license to start piloting use cases quickly
Answer: B
Enterprises are often pressured to adopt AI, and a key first step is assessing readiness with adequate governance. Buying or building before assessing skips the needed baseline. Readiness questions cover ROI, risk, ethics, skills and compliance.