AAIA
Free Advanced in AI Audit Practice Test
12 exam-style questions with full explanations — no sign-up. Score yourself, then close your gaps with the full course.
Exam fee ~$5993 exam domainsLevel Advanced2 timed practice tests in the course
✅ Free practice — no sign-up📝 Real exam-style questions💡 Detailed explanations💸 30-day money-back via Udemy
An AI audit team is reviewing a credit-scoring model's data pipeline and needs to trace which raw data sources, transformations, and feature engineering steps produced a specific training dataset version used in production. Which control provides the BEST evidence for this?
Data lineage tracking captures the full provenance chain from raw source data through each transformation and feature engineering step to the final dataset version, providing the traceability an audit requires.
A company deploys an AI system that automatically screens and ranks candidate CVs to shortlist applicants for interviews. Under the EU AI Act, how is this system classified?
Annex III explicitly lists AI systems used for recruitment or selection of natural persons, including CV screening/ranking, as high-risk, triggering Chapter 2 obligations such as risk management, data governance, human oversight, and conformity assessment.
An AI audit team is planning an audit of a fraud-detection system that calls a third-party foundation model API for real-time anomaly scoring. Which approach BEST defines the audit scope for the third-party component?
Proper scoping of third-party AI components requires leveraging contractual audit rights, obtaining vendor assurance artifacts, and testing the controls governing the integration point, consistent with AAIA guidance on third-party and model provenance assurance.
A model classifies fraudulent transactions where only 0.5% of transactions are fraudulent. Which evaluation metric should the audit team prioritize verifying was used to assess model performance, rather than relying on overall accuracy?
With severe class imbalance, precision-recall based metrics (F1, AUC-PR) properly reflect the model's ability to detect the minority fraud class, which accuracy cannot capture.
Which NIST AI RMF core function is primarily concerned with cultivating organizational culture, policies, and processes for managing AI risk, and cuts across the other three functions throughout the AI lifecycle?
Govern is the cross-cutting function establishing organizational culture, accountability structures, and policies that inform how Map, Measure, and Manage are carried out across the AI lifecycle.
An auditor must test a credit-scoring model for disparate impact across protected classes but has no access to the model's internal architecture (black-box constraint). Which technique is MOST appropriate?
Outcome-based fairness metrics computed from input/output test data are the appropriate black-box technique when internal model architecture is unavailable, directly measuring disparate treatment or impact.
During an audit of an organization's MLOps pipeline, the auditor wants to confirm that a specific production model version can be exactly reproduced if an incident requires rollback. Which artifact set MUST the model registry capture to support this?
Full reproducibility requires linking the model artifact to the exact code version, hyperparameters, and training data version used to produce it, so the model can be rebuilt or rolled back with confidence.
An organization wants an internationally recognized, third-party certifiable management system standard to demonstrate systematic governance of its AI systems to auditors and customers. Which standard should it adopt?
ISO/IEC 42001 is the AI Management System (AIMS) standard, structured similarly to ISO/IEC 27001, and is the only option among these that organizations can be certified against by an accredited third party.
During an AI audit, the auditor requests evidence of model provenance, including training data sources, intended use, limitations, and performance across demographic subgroups. Which artifact provides the MOST direct evidence of this?
A model card is the standardized documentation artifact that records training data provenance, intended use, known limitations, and subgroup performance metrics, making it primary audit evidence for model provenance testing.
A bank plans to deploy a new version of its loan-approval model that shows improved accuracy in offline testing but has not been validated against real-world traffic patterns. Which deployment approach BEST allows the audit team to verify performance risk is minimized before full rollout?
Shadow or canary deployment allows the new model's outputs to be compared against the incumbent model on real production traffic at limited exposure, surfacing real-world performance issues before full rollout.
Applying the Three Lines Model to AI governance, which party holds primary accountability for the day-to-day identification and management of risk in a specific production AI model?
In the Three Lines Model, first-line operational management — the business unit or model owner — owns and manages risk directly as part of day-to-day operations, including monitoring model performance and control effectiveness.
An auditor is evaluating a deep neural network used for medical diagnosis recommendations. The architecture is too complex to interpret directly. Which technique would BEST provide auditable evidence of which input features drove a specific prediction?
SHAP is a post-hoc, model-agnostic explainability technique that attributes prediction outcomes to specific input features, providing auditable evidence for black-box models where direct interpretation is infeasible.
Ready for the real thing?
The full course has two full-length practice tests, video lessons for every exam domain, hands-on labs and detailed answer explanations.
Start my full course on Udemy →AAIA exam — quick answers
How much does the AAIA exam cost?
The exam fee is approximately $599 and varies by region — confirm current pricing with the certification vendor before you book.
What topics are on the exam?
It covers 3 domains: AI Operations (46%), AI Governance and Risk (33%), AI Auditing Tools and Techniques (21%). 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.
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.