AIGP
Free IAPP AI Governance Professional (AIGP) Practice Test
12 exam-style questions with full explanations — no sign-up. Score yourself, then close your gaps with the full course.
Exam fee ~$6495 exam domainsLevel Intermediate2 timed practice tests in the course
✅ Free practice — no sign-up📝 Real exam-style questions💡 Detailed explanations💸 30-day money-back via Udemy
A healthcare organization is developing an AI system to assist radiologists in detecting anomalies in X-ray images. The system will analyze images and flag potential areas of concern for human review. During development, the team discovers that the training dataset contains 85% images from patients of European descent and only 15% from other ethnic backgrounds. Which action should the organization prioritize FIRST to address potential AI system risks?
Addressing dataset representativeness and bias BEFORE deployment is a foundational principle of responsible AI development. The AIGP framework emphasizes proactive risk mitigation during the development phase. Expanding the dataset to include representative samples addresses the root cause of potential bias and ensures the system performs equitably across populations before it can cause harm.
A healthcare AI system is being developed to predict patient readmission risk. During the data preparation phase, the team discovers that 78% of training records are from patients over age 65, while the hospital serves a diverse age population. The model performs well on test data but shows significant accuracy drops for patients under 40 in production. What is the MOST effective approach to address this issue?
This addresses the root cause—sample bias in training data. Collecting representative data and using stratified sampling ensures the model learns patterns across all demographic groups. This is a fundamental data quality issue that requires fixing the training distribution, which is a core practice in the AI development lifecycle for ensuring fairness and generalizability.
A healthcare organization is deploying an AI system to assist radiologists in detecting potential tumors in medical images. The system will flag images for human review but will not make final diagnoses. During testing, the model shows 92% accuracy overall, but only 78% accuracy for images from patients with darker skin tones. What is the MOST critical responsible AI principle being violated?
The scenario describes disparate performance across demographic groups (92% vs 78% accuracy), which is a clear fairness issue. Fairness in AI requires that systems perform equitably across different populations and do not systematically disadvantage particular groups. A 14-percentage-point gap in accuracy based on skin tone represents algorithmic bias that could lead to health disparities.
A financial services company is implementing an AI-powered credit scoring system that will process applications from customers across multiple countries. The system uses machine learning to assess creditworthiness. During initial testing, the risk team identifies that the model shows disparate impact against certain demographic groups. Which risk mitigation approach should be implemented FIRST to address this issue while maintaining model performance?
Fairness-aware preprocessing addresses the root cause by correcting data imbalances before they propagate through the model. Implementing fairness constraints during retraining ensures the model optimizes for both accuracy and fairness metrics simultaneously. This is a proactive approach that fixes the underlying issue rather than monitoring or compensating for it downstream.
A multinational corporation is deploying an AI-powered recruitment system that will process applicant data from candidates in the EU, UK, and California. The system uses automated decision-making to rank candidates. Which combination of regulatory frameworks must the organization primarily comply with?
Each jurisdiction has applicable laws: GDPR Article 22 restricts automated decision-making for EU residents; UK GDPR has parallel provisions post-Brexit; California's CPRA (effective 2023) includes ADMT (Automated Decision-Making Technology) provisions requiring opt-outs and disclosures. Organizations must comply with all applicable jurisdictional requirements.
Which statement BEST describes the difference between Machine Learning and Deep Learning in the context of AI system architectures?
This accurately defines the relationship between ML and DL. Deep Learning is indeed a subset of Machine Learning characterized by neural networks with multiple hidden layers that can automatically learn feature hierarchies without manual feature engineering. This is a foundational concept for understanding AI system architectures.
An AI development team must choose a model validation approach for a loan approval system that will be deployed in multiple countries with different regulatory requirements. The model will be updated quarterly. Which validation strategy BEST balances regulatory compliance and operational efficiency?
This approach combines rigorous development validation (k-fold cross-validation) with untainted holdout testing and region-specific monitoring. The separate holdout set ensures unbiased performance estimation, while continuous monitoring with region-specific metrics addresses varying regulatory requirements across jurisdictions. This aligns with best practices for regulated AI systems.
A financial services company plans to use a generative AI chatbot to provide investment advice to customers. The chatbot will recommend specific stocks and portfolio allocations. Which combination of responsible AI measures is MOST essential before deployment?
Investment advice is high-stakes with significant financial consequences. Responsible deployment requires: (1) human oversight for consequential decisions (accountability), (2) fairness testing to ensure equitable service across demographics, and (3) transparency through clear disclosures about AI capabilities and limitations. These measures address the core responsible AI principles most relevant to high-risk financial applications.
An e-commerce platform uses a generative AI system to create product descriptions. The governance team needs to establish ongoing monitoring for this system. Which combination of monitoring metrics is MOST appropriate for detecting risks specific to generative AI outputs?
These metrics directly address generative AI-specific risks: hallucinations (factuality), harmful content generation (toxicity), security vulnerabilities (prompt injection), and regulatory compliance (policy violations). This monitoring approach detects the unique failure modes of large language models and generative systems.
An organization subject to the EU AI Act is developing a credit scoring system. Under the AI Act's risk classification framework, what obligations does this create?
Credit scoring and creditworthiness evaluation are explicitly listed in Annex III of the EU AI Act as high-risk AI systems (banking and financial services category). High-risk systems must meet strict requirements including conformity assessments, CE marking, quality management systems, record-keeping, transparency, human oversight, accuracy, robustness, and cybersecurity measures before deployment.
A financial services company is implementing an AI-based credit scoring system. During testing, they discover that the model's accuracy is 92% overall, but when analyzed by demographic groups, accuracy is 95% for one group and 85% for another group. The governance team must decide on next steps. According to AI governance best practices, what should be their PRIMARY concern?
Disparate performance across demographic groups is a critical fairness concern in AI governance. A 10-percentage-point accuracy gap between groups indicates the system may produce discriminatory outcomes. AI governance frameworks require identifying and mitigating such disparities before deployment, especially in high-stakes domains like financial services where regulatory requirements (e.g., fair lending laws) apply.
During the model development phase, a team building a content moderation AI discovers their precision is 0.91 but recall is 0.64 for detecting policy violations. The business stakeholder states that failing to catch violations (false negatives) creates significant legal and brand risk. What is the MOST appropriate action?
Given the business requirement that false negatives pose significant risk, increasing recall (catching more violations) is the priority. Adjusting the decision threshold is a direct, effective method to shift the precision-recall trade-off. Pairing this with human review addresses the expected increase in false positives (lower precision), creating a practical solution that aligns model performance with business risk tolerance.
Ready for the real thing?
The full course: two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed explanations.
$109.99 $34.99 with code FREETEST33 — valid through August 22.
Get my $34.99 deal →AIGP exam — quick answers
How much does the AIGP exam cost?
The exam fee is approximately $649 and varies by region — confirm current pricing with the certification vendor before you book.
What topics are on the exam?
It covers 5 domains: Foundations of AI Systems (~13%), AI Development Lifecycle (~15%), Responsible AI Principles and Impacts (~20%), AI Risk Management and Governance Implementation (~25%), AI Laws, Regulations, and Standards (~27%). The full course has a dedicated chapter, lab and practice-test coverage for each.
Is this practice test really free?
Yes — all questions on this page are free with explanations and no sign-up. The paid Udemy course adds two full-length timed exams, video lessons and hands-on labs.
How do I get the discount?
Use code FREETEST33 at checkout for $34.99 (list $109.99) through August 22 — the enroll button applies it automatically.
Will this prepare me for the real exam?
The questions mirror the real exam's style and are mapped to the official domains. This is exam-focused preparation — combine the free test with the full course's timed simulations to gauge your readiness.