AI0-001 AI Security, Ethics and Governance Practice Question
A hospital deploys an AI model to predict patient readmission risk. The compliance team asks which TWO technical controls help comply with data minimization principles under AI governance frameworks. (Choose two.)
⚠ Common exam trap
The trap here is assuming that any privacy-related measure, such as encryption or consent, automatically fulfills data minimization requirements.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Use federated learning to keep raw data on local devices.
Differential privacy and federated learning both limit how much individual data influences the model or leaves its source, which is the core of data minimization. Encryption, consent, and simple de-identification do not reduce the scope of data collection or use, so they do not satisfy the principle as directly.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Use federated learning to keep raw data on local devices.
Why this is correct
Federated learning trains a shared model by exchanging only model updates, not raw patient records, so sensitive data never leaves the local environment. This reduces the volume of data centralized, satisfying data minimization by design. It is particularly relevant in healthcare, where moving patient data across systems increases regulatory and breach risk.
- ✓
Apply differential privacy during model training.
Why this is correct
Differential privacy adds calibrated noise to training data or gradients, so the model's outputs reveal little about any single patient. This directly supports data minimization by limiting the influence of individual records, making it harder to infer specific patient details from the model. It is a recognized technical control for privacy-preserving AI and aligns with governance requirements for reducing data exposure.
- ✗
Anonymize data by removing patient names before training.
Why it's wrong here
Removing names is a basic de-identification step, but it is insufficient for data minimization because other fields can still re-identify individuals. It does not limit the amount of data collected or the model's ability to memorize sensitive attributes. Stronger technical controls are needed to meet governance expectations.
- ✗
Store all training data in a single encrypted data lake.
Why it's wrong here
Encryption protects data at rest but does not reduce the amount of data collected or retained. Data minimization is about limiting collection and use, not just securing storage. A single data lake may actually increase exposure by centralizing sensitive records, so this control does not satisfy the minimization principle.
- ✗
Require users to consent to broad data usage terms.
Why it's wrong here
Consent is a legal and ethical requirement, but broad consent does not minimize data collection or use. It may even permit more extensive processing. Data minimization requires technical and procedural limits on what data is collected and how it is used, not just obtaining permission for broad usage.
About these practice questions
One of 962 original AI0-001 practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official CompTIA exam blueprint
This AI0-001 practice question is part of Courseiva's free CompTIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI0-001 exam.