AIF-C01 Fundamentals of Generative AI Practice Question

A financial services firm is deploying a generative AI assistant that answers employee questions about internal policies. Compliance requires that every response cite the exact policy document and section used. The assistant currently relies only on the foundation model's pretrained knowledge and frequently invents policy details. Which technique should the team implement to ground responses in the firm's own documents and produce citations?

⚠ Common exam trap

The trap here is believing that fine-tuning on domain text guarantees accurate citations, when only retrieval of the actual source documents provides verifiable provenance.

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

✓

Implement Retrieval Augmented Generation by embedding the policy documents in a vector store and retrieving relevant passages at query time.

Grounding a model in proprietary content requires supplying that content at inference time rather than relying on pretrained weights. Retrieval Augmented Generation retrieves the most relevant document chunks and places them in the prompt, so the model's answer is conditioned on real policy text and can reference the source document and section. Sampling or length adjustments do not add knowledge.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Implement Retrieval Augmented Generation by embedding the policy documents in a vector store and retrieving relevant passages at query time.
Why this is correct
Retrieval Augmented Generation embeds source documents, retrieves the passages most semantically similar to the user's question, and passes them into the prompt so the model answers from that supplied context. Because the retrieved chunks carry document and section metadata, the assistant can cite the exact source. It also keeps answers current as policies change without retraining.
✗
Increase the model's temperature setting so it explores more of its pretrained knowledge about corporate policy.
Why it's wrong here
Temperature controls randomness in token sampling. Raising it makes output more varied and creative, which increases the likelihood of fabricated policy details rather than reducing them. It does nothing to inject the firm's actual documents into the model's context, so citations to specific sections could not be produced reliably. This directly worsens the compliance problem.
✗
Lower the maximum token limit so the assistant gives shorter answers that are less likely to contain errors.
Why it's wrong here
Shortening responses reduces verbosity but does not supply the model with the firm's policies, so hallucinations can still occur in a single sentence. It also removes no risk of fabricated section numbers. Token limits address cost and latency, not factual grounding, so this does not satisfy the compliance requirement for citations.
✗
Fine-tune the foundation model on a large corpus of public financial regulations.
Why it's wrong here
Fine-tuning on public regulations changes the model's weights toward general regulatory language but does not give it access to the firm's internal policy documents, nor does it guarantee verbatim citations. It is also costly and slow to refresh when policies change. The assistant would still lack provenance for its answers, failing the citation requirement.

Go deeper

AIF-C01 Fundamentals of Generative AI Practice Question

Why each option matters

Option-by-option breakdown

Related to this question