CSA TAISE - Trusted AI Safety Expert v1 Exam
Page: 3 / 12
Total 60 questions
Question #11 (Topic: Exam A)
Sampling bias in AI occurs when:
A. Historical patterns in data reflect past discrimination
B. The AI system is used outside its intended purpose
C. Training data is not representative of the real-world population the AI system will serve
D. Evaluation methods use inappropriate benchmarks for testing
Answer: C
Question #12 (Topic: Exam A)
A credit risk model was trained using financial data from five years ago. What data quality problem might cause it to make poor predictions about customers today?
A. The historical data contains too many missing values
B. The data was collected using inconsistent methodologies
C. The data lacks timeliness for the current economic context
D. The financial features are not relevant for credit decisions
Answer: C
Question #13 (Topic: Exam A)
What is the relationship between model explainability and AI trustworthiness?
A. They are unrelated concepts
B. Explainability only matters for generative AI
C. Trustworthiness is more important than explainability
D. Explainability is essential for evaluating trustworthiness
Answer: D
Question #14 (Topic: Exam A)
When would you choose LIME over SHAP for model explanation?
A. When computational resources are limited and you need quick local explanations
B. When explaining models to non-technical audiences who need visual simplicity
C. When stakeholders need easily interpretable approximations rather than exact feature contributions
D. When you need to explain highly complex ensemble models with many features
Answer: A
Question #15 (Topic: Exam A)
The 2021 Anthony Bourdain documentary deepfake controversy highlighted:
A. Legal issues with documentary filmmaking standards
B. The ethical implications of using AI to create voiceovers without someone's consent
C. Technical limitations in voice synthesis technology
D. Copyright violations in using someone's voice without permission
Answer: B
