CCAO-F Claude Model Fundamentals Practice Question
A developer is integrating Claude into a legal document review system. The system must process lengthy contracts and answer questions about specific clauses. Which TWO of the following techniques help mitigate the risk of Claude hallucinating details not present in the document? (Choose two.)
⚠ Common exam trap
The trap here is thinking that adjusting generation parameters like temperature or max_tokens can prevent hallucinations, when grounding instructions and citations are the key.
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
✓
Instruct Claude to answer only based on the provided document and to say 'I don't know' if the information is not present.
To reduce hallucinations when reviewing legal documents, the most effective methods are to instruct Claude to rely solely on the provided text and to require citations for its answers. These techniques ground the model's responses in the source material, making it less likely to invent details and easier to verify accuracy.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Use a higher temperature setting to encourage more creative responses.
Why it's wrong here
Higher temperature increases randomness and creativity, which is counterproductive for factual tasks like legal document review. It would make Claude more likely to invent details or deviate from the provided text. For accuracy and grounding, a lower temperature (e.g., 0) is recommended to make outputs more deterministic and focused on the context.
- ✓
Instruct Claude to answer only based on the provided document and to say 'I don't know' if the information is not present.
Why this is correct
Explicitly instructing Claude to ground its answers in the provided document and to admit uncertainty when information is missing reduces hallucinations. This prompt engineering technique encourages the model to rely on the context rather than its internal knowledge, which is crucial for legal documents where accuracy is paramount. It sets clear boundaries for the model's responses.
- ✗
Set max_tokens to a very low value to limit the response length.
Why it's wrong here
Limiting response length does not address hallucinations; it only truncates the output. A truncated answer could be incomplete or misleading, and the model might still include fabricated details within the allowed tokens. This technique is unrelated to improving factual accuracy or grounding in the provided document.
- ✓
Provide the full contract text in the prompt and ask Claude to cite the specific section for each answer.
Why this is correct
Including the full contract and requiring citations forces Claude to reference the source material, reducing the likelihood of fabricating information. This technique, known as citation grounding, makes it easier to verify answers and holds the model accountable to the provided text. It leverages Claude's ability to handle long contexts and extract relevant passages.
- ✗
Fine-tune the model on a dataset of legal contracts.
Why it's wrong here
Fine-tuning is not available for all Claude models and requires significant resources. Moreover, fine-tuning on legal contracts might improve general legal knowledge but does not directly prevent hallucinations about a specific document provided in the prompt. For mitigating hallucinations in a given context, prompt-based grounding techniques are more immediate and effective.
About these practice questions
This CCAO-F question is part of Courseiva's 259-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam 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 Anthropic exam blueprint
This CCAO-F practice question is part of Courseiva's free Anthropic 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 CCAO-F exam.