AI0-001 AI Security Practice Question
A machine learning team is developing a model to predict loan defaults using sensitive customer financial data. They need to share the model with third-party auditors without exposing individual customer records. Which privacy-preserving technique allows auditors to query the model while providing mathematical guarantees about the privacy of the training data?
⚠ Common exam trap
A common misconception is that federated learning inherently provides privacy guarantees, when in fact it only addresses data locality and does not prevent model inversion or membership inference attacks without additional differential privacy mechanisms.
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
✓
Differential privacy
Differential privacy is correct because it adds calibrated noise to the model's training process or query responses, providing a formal mathematical guarantee (ε-differential privacy) that the inclusion or exclusion of any single individual's data does not significantly affect the output. This allows auditors to query the model without exposing individual customer records, as the noise bounds the information leakage from the training data.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Differential privacy
Why this is correct
Differential privacy provides a formal epsilon guarantee bounding how much any single customer record changes query outputs, letting auditors query the model without exposing individual records. This satisfies the demand for mathematical privacy guarantees on the training data.
- ✗
Federated learning
Why it's wrong here
Federated learning trains a shared model across distributed data holders without centralising records, but it does not let auditors query an already-trained model with provable privacy. It is tempting because it keeps raw data local, which suits training across hospitals or banks that cannot pool data.
- ✗
k-anonymity
Why it's wrong here
k-anonymity generalises quasi-identifiers in a released dataset, so it protects published records rather than answering queries against a trained model, and it offers no formal privacy guarantee. It is tempting because it is a recognised anonymisation technique, correct when releasing tabular data for analysis.
- ✗
Homomorphic encryption
Why it's wrong here
Homomorphic encryption allows computation on ciphertext, but it protects data during processing rather than providing a mathematical privacy guarantee for the training data behind model queries. It is tempting because it enables secure outsourced computation, correct when a third party must process encrypted inputs.
About these practice questions
This AI0-001 question is part of Courseiva's 962-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
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.