AI0-001 AI Security, Ethics and Governance Practice Question
Which TWO of the following are common methods for mitigating bias in AI models?
⚠ Common exam trap
CompTIA often tests the distinction between bias mitigation techniques (pre-processing, in-processing, post-processing) and general ML best practices like regularization or cross-validation, leading candidates to confuse L1 regularization or k-fold cross-validation with fairness methods.
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
✓
Reweighting training samples based on sensitive attributes
Option B (Reweighting training samples based on sensitive attributes) is correct because it is a standard pre-processing bias-mitigation technique: by assigning higher weights to underrepresented or historically disadvantaged groups, the model's loss function is adjusted so those samples contribute more to the learned parameters, reducing disparate impact across sensitive attributes. Option D (Adding fairness constraints during training) is correct because it is a standard in-processing technique: fairness metrics such as demographic parity, equalized odds, or disparate impact are encoded as constraints or penalty terms in the objective function, forcing the optimizer to trade off accuracy against a quantified fairness criterion. The other options do not belong: A (adversarial training) is primarily used to improve robustness against adversarial examples, not to mitigate bias; C (L1 regularization) induces sparsity in weights for feature selection and generalization, not fairness; and E (k-fold cross-validation) is a model-evaluation/resampling method for estimating generalization performance, not a bias-mitigation method.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Using adversarial training
Why it's wrong here
Adversarial training hardens a model against perturbed inputs to improve robustness, not fairness; it does not address skewed sampling or protected-attribute imbalance in the training set. It is tempting because it sounds like it targets model weaknesses, and it would be correct when defending against adversarial examples or evasion attacks.
- ✓
Reweighting training samples based on sensitive attributes
Why this is correct
Reweighting assigns higher weights to under-represented samples during training, directly countering the skewed class distributions that produce biased predictions. This satisfies the stem's mitigation requirement by adjusting the model's learned decision boundary rather than merely auditing outcomes post hoc, making it a recognised pre-processing bias mitigation technique.
- ✗
Applying L1 regularization
Why it's wrong here
L1 regularization shrinks weights to produce sparse models, addressing overfitting rather than bias; it does not correct skewed training data or imbalanced class representation. It is tempting because it is a genuine model-tuning technique, and it would be the right choice when the goal is feature selection or reducing model complexity.
- ✓
Adding fairness constraints during training
Why this is correct
Fairness constraints encode parity or equalised-odds conditions directly into the training objective, penalising disparate outcomes across protected groups. This addresses the stem's requirement to mitigate bias at its source, rather than only detecting it post hoc through auditing or documentation.
- ✗
Performing k-fold cross-validation
Why it's wrong here
k-fold cross-validation estimates generalisation performance by partitioning data into folds; it does not alter training data composition or reweight under-represented groups, so bias persists. It is tempting because it is a standard evaluation practise, and it would be correct when you need a reliable estimate of model accuracy on unseen data.
Quick reference
RAID Level Comparison
| RAID Level | Min Disks | Fault Tolerance | Read | Write | Usable Capacity |
|---|---|---|---|---|---|
| RAID 0 | 2 | None | Excellent | Excellent | 100% |
| RAID 1 | 2 | 1 disk | Good | Moderate | 50% |
| RAID 5 | 3 | 1 disk | Good | Moderate | 67–94% |
| RAID 6 | 4 | 2 disks | Good | Lower | 50–88% |
| RAID 10 | 4 | 1 disk per mirror | Excellent | Good | 50% |
RAID is not a backup strategy — it protects against disk failure but not against accidental deletion, ransomware, or site-level events.
About these practice questions
Courseiva writes every AI0-001 question from scratch — 962 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or 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.