Prompt
source attribution in AI answers
Latest observation
Nov 18, 2026
OpenAI APIWeb search: off
When giving answers, source attribution is important because it helps users verify claims, assess reliability, and distinguish between model-generated inference and information drawn from specific references.
A good source attribution practice in AI answers usually includes:
- Identify the source clearly: name the author, organization, paper, website, dataset, or document.
- Distinguish fact from inference: indicate whether something is directly stated in a source or inferred from it.
- Be specific: cite the exact section, page, or URL if available.
- Be current when needed: for time-sensitive topics, note the date or version of the source.
- Avoid fabricated citations: never invent references or imply you consulted sources you did not.
- Use uncertainty honestly: if evidence is weak or conflicting, say so.
Example:
- Good: “According to the WHO fact sheet on diabetes (2023), type 2 diabetes is the most common form.”
- Better for nuance: “The WHO fact sheet says type 2 diabetes is the most common form; this conclusion is from their 2023 summary, not a clinical guideline.”
If you want, I can also give you:
- a template for citing sources in AI responses,
- best practices for RAG/citations, or
- a policy-style rule set for source attribution.
Brands