AIF-C01 Practice Question: Security, Compliance, and Governance for AI Solutions
A company uses Amazon Bedrock and wants to ensure that the model outputs are grounded in a set of provided documents to reduce hallucinations. Which TWO actions should they take? (Select TWO.)
⚠ Common exam trap
The trap is confusing fine-tuning with RAG. Fine-tuning adjusts model weights but does not provide a mechanism to ground responses in specific documents at runtime. Also, word filters are often mistaken for grounding mechanisms, but they only block specific terms, not verify factual consistency.
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
✓
Enable the grounding check in Bedrock Guardrails
Option A is correct because Bedrock Guardrails provides a contextual grounding check that evaluates model responses against a provided source (reference) and applies a grounding threshold to filter or flag responses that are not supported by the source, directly reducing hallucinations. Option D is correct because Amazon Bedrock Knowledge Bases ingests the provided documents, chunks them, generates embeddings, and stores them in a vector store so the model can retrieve relevant chunks at inference time via RetrieveAndGenerate, grounding outputs in the actual documents. Option B is wrong because a word filter blocks specific words or phrases and cannot determine whether a statement is factually grounded in the source documents. Option C is wrong because model invocation logging to S3 only records requests and responses for auditing and monitoring; it does not ground outputs or prevent hallucinations. Option E is wrong because fine-tuning on the documents adapts model weights to style and patterns but does not provide retrieval-based grounding or verifiable citations, and it is not the recommended mechanism for grounding against a document set.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Enable the grounding check in Bedrock Guardrails
Why this is correct
The grounding check compares each model response against the retrieved source chunks and flags or blocks claims unsupported by them, directly reducing hallucination. It satisfies the grounding requirement by validating output against provided documents rather than relying on the model's parametric knowledge.
- ✗
Configure a word filter to block ungrounded phrases
Why it's wrong here
A word filter blocks output patterns; it cannot verify whether generated text is supported by the supplied documents. Filters suit content moderation or profanity blocking. Grounding requires retrieval-augmented generation, where the documents are retrieved and passed to the model as context.
- ✗
Enable model invocation logging to S3
Why it's wrong here
Invocation logging records prompts and responses to Amazon S3 for auditing and monitoring; it does not influence what the model generates. Logging suits compliance and troubleshooting. Grounding requires retrieval-augmented generation so the supplied documents are retrieved and passed as context.
- ✓
Use Amazon Bedrock Knowledge Bases to store and retrieve document chunks
Why this is correct
Knowledge Bases performs chunking, embedding and vector storage of the documents, then retrieves the most relevant chunks at query time for inclusion in the prompt. This supplies the source material the grounding check validates against, satisfying the requirement to ground outputs in provided documents.
- ✗
Fine-tune the model on the documents
Why it's wrong here
Fine-tuning adjusts model weights for style, tone or task behaviour; it does not cite or constrain output to specific supplied documents at inference time. Grounding needs retrieval-augmented generation, which injects the current documents into the prompt as context.
Quick reference
AWS S3 Storage Class Comparison
| Storage Class | Min Duration | Retrieval | Use Case |
|---|---|---|---|
| S3 Standard | None | Immediate | Frequently accessed data |
| S3 Standard-IA | 30 days | Immediate | Infrequent access, rapid retrieval |
| S3 One Zone-IA | 30 days | Immediate | Non-critical infrequent data |
| S3 Intelligent-Tiering | None | Immediate–hours | Unknown or changing access patterns |
| S3 Glacier Instant | 90 days | Milliseconds | Archive with instant retrieval |
| S3 Glacier Flexible | 90 days | Minutes–hours | Archive, flexible retrieval |
| S3 Glacier Deep Archive | 180 days | Hours | Long-term compliance archive |
Go deeper
Related to this question
About these practice questions
Courseiva writes every AIF-C01 question from scratch — 862 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 and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Amazon Web Services exam blueprint
This AIF-C01 practice question is part of Courseiva's free Amazon Web Services 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 AIF-C01 exam.