Security, Compliance, and Governance for AI Solutions
This domain carries 14% of the AIF-C01 exam. Practise it with original questions that explain every answer option.
- Certification
- AIF-C01
- Exam weight
- 14%
- Subdomains
- 2
- Objectives
- 12
What this domain covers
Example question from this domain
A healthcare assistant gives a confident medical answer containing a claim that is unsupported by its approved sources. Which failure is illustrated?
- A hallucination that grounding or output validation should detect
- Toxicity caused by abusive language
- Demographic bias caused by unequal subgroup performance
- Prompt injection that changes the system instructions
A hallucination that grounding or output validation should detect — A confident but unsupported claim is a hallucination; grounding and output validation can reduce or detect it.
Toxicity caused by abusive language — The issue is factual support, not abusive or toxic language.
Demographic bias caused by unequal subgroup performance — The scenario does not describe a performance disparity between demographic groups.
Prompt injection that changes the system instructions — No adversarial instruction or attempt to override the system prompt is described.
Continue with AIF-C01
Domain names and weightings follow the vendor's published exam guide. ExamNova questions are written in-house and are not exam dumps.