AAIR Practice Test 2 — 30 Questions

Question 1

Why is 'AI Literacy' for the board of directors a key component of AI Governance?

Accepted Answer

To enable effective decision-making regarding AI adoption and risk management.. Directors must understand AI risks and benefits to provide informed oversight and define the organization's risk appetite.

Answer

To allow directors to code their own AI models.

Answer

To replace the need for the Chief Risk Officer.

Answer

To reduce the cost of IT infrastructure.

Question 2

The AI Risk Governance Committee is defining the risk appetite for a new generative AI chatbot. Which THREE factors must be considered to align with organizational risk tolerance?

Accepted Answer

Data privacy and residency requirements for training sets. Option B is correct because data privacy and residency requirements directly shape risk appetite: regulations like GDPR, HIPAA, or data-localization laws determine what training data may be used and where it can be stored, which bounds the acceptable risk for the chatbot. Option C is correct because model drift is a core AI risk factor — as data distributions shift, output accuracy degrades, so the committee must weigh the likelihood of drift against the organization's tolerance for degraded performance. Option D is correct because hallucination is a generative-AI-specific risk whose potential impact on brand reputation and customer trust must be assessed against the organization's risk tolerance. Option A is not a risk-appetite factor; sprint cadence is a delivery-process metric, not a measure of AI risk exposure. Option E is also not a risk-tolerance factor; GPU infrastructure cost is a budgeting concern, not a determinant of acceptable risk.

Answer

The speed of the development team's sprint cycles

Answer

The total cost of GPU infrastructure

Question 3

When defining KPIs for an AI Governance program, which metric is most predictive of long-term model robustness?

Accepted Answer

The frequency of model retraining triggered by performance drift alerts.. The frequency of model retraining triggered by performance drift alerts is the most predictive metric of long-term robustness because it reflects the organization's ability to detect degradation and respond with corrective action. A governance program that tracks this metric demonstrates closed-loop monitoring and remediation, which sustains model performance over time as data distributions shift. This aligns with MLOps best practices for drift detection and continuous improvement.

Answer

Number of users accessing the AI interface.

Answer

Total number of AI models currently in production.

Answer

The budget spent on GPU cloud computing resources.

Question 4

A company is integrating an AI model that uses 'Online Inference'. What is the most significant risk regarding model security?

Accepted Answer

The risk of model inversion attacks on the inference endpoint.. Online inference endpoints are exposed to the public/network, creating surface area for adversarial attacks.

Answer

The risk of the training data being corrupted by the user.

Answer

The risk of high database storage costs.

Answer

The risk of the Python interpreter crashing.

Question 5

A practitioner is managing AI risk in a CI/CD pipeline. What should be the final gate before model promotion to production?

Accepted Answer

Automated performance validation against a hold-out test set with defined threshold gates.. Automated validation of performance metrics against a 'golden dataset' ensures the model meets the required threshold.

Answer

Increasing the memory of the production server.

Answer

Updating the project's documentation.

Answer

A peer review of the training code.

Question 6

In the context of the AI RMF, what is the significance of the 'Measure' function for risk management?

Accepted Answer

To evaluate the model's performance and risk outcomes against established benchmarks.. The Measure function of the AI RMF is responsible for evaluating AI systems' performance, trustworthiness characteristics, and risk outcomes against defined benchmarks and metrics. It provides the quantitative and qualitative evidence needed to determine whether risks are within tolerance and whether the system meets its intended purpose. This function operationalizes the risk assessment by turning identified risks into measurable indicators.

Answer

To identify which employees have access to the AI system.

Answer

To set the budget for future AI projects.

Answer

To provide a score for the developers' performance.

Question 7

Which of the following is a classic example of 'AI Hallucination' that needs monitoring?

Accepted Answer

The system generates confident but factually incorrect information. AI hallucination refers to a model generating confident but factually incorrect or nonsensical information, which is a key risk that requires monitoring. This can lead to misinformation, user harm, and reputational damage. The other options describe operational or infrastructure issues, not hallucination.

Answer

The system is running slowly

Answer

The system requires too much electricity

Answer

The system password has expired

Question 8

Which TWO metrics are most appropriate for measuring the effectiveness of an AI Risk Program?

Accepted Answer

Total number of AI-related incidents reported and remediated.. Option D is correct because tracking the total number of AI-related incidents reported and remediated directly measures how well the AI Risk Program detects, responds to, and resolves real risk events, which is a core effectiveness indicator. Option E is correct because the percentage of AI models that have undergone a formal risk assessment measures the program's coverage and consistency in applying risk controls across the AI model inventory, reflecting proactive risk management maturity. Options A, B, and C are not appropriate because average data scientist salary, coffee machine count, and social media followers are unrelated to AI risk identification, assessment, mitigation, or governance outcomes and do not indicate program effectiveness.

Answer

The average salary of data scientists.

Answer

Number of coffee machines in the building.

Answer

Number of social media followers of the company.

Question 9

Which THREE strategies are effective for mitigating 'Adversarial Risk' in an AI lifecycle?

Accepted Answer

Rate limiting on API calls to prevent brute-force querying.. Option B is correct because rate limiting on API calls throttles the query volume an attacker can send, which directly counters brute-force and model-extraction attacks that rely on high-volume probing of the inference endpoint. Option D is correct because adversarial training explicitly augments the training set with perturbed examples, hardening the model's decision boundaries against evasion and adversarial-input attacks. Option E is correct because input sanitization and validation on all model endpoints rejects malformed, out-of-distribution, or injection-style payloads before they reach the model, reducing the attack surface for adversarial manipulation. Option A does not belong because adding documentation staff addresses process and knowledge management, not adversarial manipulation of the model. Option C does not belong because a more expensive cloud provider changes cost, not the security posture against adversarial inputs.

Answer

Hiring more documentation staff.

Answer

Using a more expensive cloud provider.

Question 10

When addressing 'Explainability' as an AI risk, which approach is most effective for a non-technical stakeholder?

Accepted Answer

Providing counterfactual explanations like 'If your income had been $5k higher, the loan would have been approved'.. Counterfactual explanations are the most effective for non-technical stakeholders because they describe what would need to change in the input for a different outcome, using plain language and concrete scenarios. For example, 'If your income had been $5k higher, the loan would have been approved' is immediately understandable without any mathematical background. This approach aligns with GDPR Article 22 and the right to meaningful information about automated decisions.

Answer

Presenting the feature weights of a neural network.

Answer

Explaining the gradient descent optimization process.

Answer

Showing the mathematical formula used for loss reduction.

Question 11

What is the first step in performing an AI Risk Assessment?

Accepted Answer

Defining the scope, context, and purpose of the AI system.. Per ISO 31000 and NIST AI RMF, every risk assessment begins with establishing scope, context, and purpose — defining which AI system is being assessed, its boundaries, stakeholders, operating environment, and the objectives of the assessment. Without this foundational step, subsequent activities like bias testing or control selection lack a reference frame and may address the wrong risks or miss critical ones. This mirrors ISO 31000 Clause 6.3.1, which explicitly places 'defining scope and context' as the first activity in the risk assessment process.

Answer

Hiring a security consultant.

Answer

Updating the company website.

Answer

Running a bias test.

Question 12

Which role is primarily responsible for ensuring that AI-generated content adheres to intellectual property risk policies?

Accepted Answer

Chief Legal Officer. The Chief Legal Officer is primarily responsible for ensuring that AI-generated content adheres to intellectual property risk policies, as they oversee legal compliance and IP strategy. This role ensures that AI outputs do not infringe on copyrights, trademarks, or other IP rights. Other roles lack the legal expertise and authority.

Answer

System Administrator

Answer

Product Manager

Answer

Data Engineer

Question 13

A board of directors requests a dashboard to monitor AI risk exposure. Which metric provides the most effective board-level oversight regarding AI regulatory compliance?

Accepted Answer

Percentage of critical AI systems that have undergone a formal regulatory compliance impact assessment.. The board's primary concern is regulatory compliance and risk exposure, not operational performance. The percentage of critical AI systems that have undergone a formal regulatory compliance impact assessment directly measures the organization's adherence to frameworks like the EU AI Act, NIST AI RMF, or ISO/IEC 42001. This metric provides assurance that high-risk AI systems are being evaluated against legal and ethical requirements, which is a core board-level oversight responsibility.

Answer

Mean Time to Detect (MTTD) for model adversarial attacks.

Answer

Current GPU utilization rates for production inference clusters.

Answer

Average latency of API requests for generative AI models.

Question 14

You are mapping existing ISO 31000 risk management processes to an AI-specific application. Which step is most crucial to ensure 'Risk Treatment' is appropriately scaled for AI?

Accepted Answer

Implementing a tiered control framework where mitigations scale based on the system's impact assessment.. A tiered control framework is the correct answer because ISO 31000's Risk Treatment step requires selecting controls proportionate to the assessed risk level. For AI, this means mitigations (e.g., human-in-the-loop review, explainability requirements, monitoring intensity) must scale with the system's impact classification — a low-risk internal chatbot needs lighter controls than a high-risk hiring or medical diagnostic model. This impact-based tiering directly operationalizes ISO 31000's principle that treatment options be evaluated against risk criteria and organizational context.

Answer

Automating all risk logging in the SIEM.

Answer

Discontinuing the use of third-party AI APIs to avoid vendor risk.

Answer

Standardizing all AI models under one security policy regardless of function.

Question 15

When an AI model is updated (e.g., retrained on new data), what action is required from an AI Risk Governance perspective?

Accepted Answer

The model should be re-validated to ensure it still meets risk thresholds.. Retraining a model on new data changes its behavior, so from a risk governance perspective it must be re-validated against the same risk thresholds (accuracy, fairness, robustness, compliance) before being trusted in production. Renaming or assuming no action is needed ignores the fact that the model's outputs and risk profile may have changed. Automatic deployment without review bypasses required oversight.

Answer

The model should be renamed to show it is a new version.

Answer

No action is needed as the underlying architecture is unchanged.

Answer

The model should be automatically deployed without review.

Question 16

Which of the following is an example of 'Data Quality Risk' in the AI lifecycle?

Accepted Answer

High rates of mislabeled training data.. High rates of mislabeled training data are a classic Data Quality Risk because labels directly determine what the model learns; systematic mislabeling introduces bias and degrades accuracy. Data quality risks encompass issues with accuracy, completeness, consistency, and provenance of training data. Mislabeling is a data integrity problem that propagates into model behavior and is hard to detect post-training.

Answer

Setting a high learning rate.

Answer

Using Python instead of R.

Answer

Using a GPU that is two generations old.

Question 17

When setting up an AI risk monitoring dashboard, what is the most important consideration for ensuring executive buy-in?

Accepted Answer

Map AI performance metrics to business impact indicators like customer churn or operational downtime.. Executive buy-in depends on demonstrating business value, not technical detail. Mapping AI performance metrics to business impact indicators such as customer churn, revenue, or operational downtime translates model behavior into outcomes executives already care about and are measured on. This makes the dashboard a decision-support tool rather than an engineering report.

Answer

Include the number of GPU hours utilized.

Answer

Display the raw training error rates.

Answer

Show a comparison of the organization's AI progress against competitors.

Question 18

When an AI model fails in production, what is the 'Post-Mortem' goal?

Accepted Answer

To identify root causes and update risk management practices. A post-mortem's goal is to identify root causes of the failure and feed those lessons back into risk management practices, controls, and monitoring. It is a blameless, learning-oriented exercise that improves future resilience rather than assigning fault. This aligns with AAIR's emphasis on continuous improvement of the AI risk management lifecycle.

Answer

To hide the failure from management

Answer

To fire the developer responsible

Answer

To delete all logs associated with the model

Question 19

Why is it critical to have 'Human-in-the-Loop' (HITL) for high-risk AI decisions?

Accepted Answer

To provide accountability and a safeguard against errors. Human-in-the-Loop (HITL) ensures that a qualified human reviews and can override high-risk AI decisions, providing both accountability (a responsible person is answerable for the outcome) and a safeguard against model errors, bias, or edge cases the model was not trained for. This is essential for high-stakes domains like healthcare, finance, and criminal justice.

Answer

To make the computer faster

Answer

To replace the need for model testing

Answer

To save money on hardware

Question 20

When establishing KPIs for an AI program, what is the risk of focusing solely on 'Model Accuracy'?

Accepted Answer

Other critical risks such as bias, security, and fairness may be overlooked.. Focusing solely on model accuracy creates a narrow optimization target that ignores other critical AI risks such as bias, fairness, security, privacy, and robustness. Accuracy alone does not guarantee that a model is ethical, compliant, or safe in production, so other risk dimensions can be neglected.

Answer

The model will be too computationally expensive.

Answer

The data scientists will take too long to develop the model.

Answer

The model will not be deployable in production.