- Duale AI
- Documentation
- Web research
- Sources and factuality
Sources and factuality in web research
This page defines web research terms like source, citation, and grounded claim, and describes how to verify material claims against fetched evidence.
This page covers factuality definitions, term distinctions, typed result checks, claim verification, freshness, and proof separation in web research.
- Factuality definitions and web research step distinctions.
- Term meanings for discovery results, sources, citations, and fact-checks.
- Typed result checks and their limits.
- Material claim verification and acceptance criteria.
- Freshness accounting and proof separation.
Summaries were generated by AI.
Here, factuality means whether each claim in an answer matches the relevant facts.
Use sources to make an answer easier to inspect, not to claim certainty. A web research workflow has several separate steps: discover candidate pages, read selected pages, connect evidence to claims, and review the conclusions. Web access, grounding, and fact-checking are not the same function.
- Term
- Discovery result
- Meaning in Web research
- A page matched a search query
- Do not infer
- The page was read, supports the claim, or is authoritative
- Term
- Source
- Meaning in Web research
- The research selected a page or document as evidence for a claim
- Do not infer
- The evidence is reliable, complete, current, or applicable
- Term
- Citation
- Meaning in Web research
- An answer refers to a source, usually with a URL
- Do not infer
- The source supports the nearby claim or the answer represents it correctly
- Term
- Grounded claim
- Meaning in Web research
- The research developed or checked a claim against fetched source material
- Do not infer
- The material or claim is true, current, complete, or independent
- Term
- Fact-check
- Meaning in Web research
- A process compares a precise claim with evidence and handles conflicts
- Do not infer
- A guarantee that the conclusion or later decision is correct
A result can contain citations without grounded claims. More than one URL also does not prove independent support: two publishers can repeat the same underlying report. A fact-check can remain unresolved when evidence conflicts or is incomplete.
Know what a typed result checks
The terminal SDK result follows your requested response model. Include the source fields your application must retain. A result that matches that model does not prove which pages were read, that an excerpt matches its source, that sources are independent, or that a claim is true. Open and check every material source.
Verify a material claim
Use your own acceptance test for a material claim:
- State the exact claim, date, and scope.
- Open the final URL and locate the material that supports the whole claim.
- Check the publisher, evidence origin, publication date, and effective date.
- For corroboration, confirm that another source has independent evidence instead of repeating the first source.
- Record conflicts and gaps, then apply the review that your decision requires.
Accept the claim only when the accessible evidence supports its full wording and scope, its dates match the decision, and your required review is complete. Otherwise, qualify the claim, report the gap, omit it, or send it to a qualified reviewer.
Account for freshness
Search can use cached results and upstream indexes. A page can change after the task. The SDK does not return a public content snapshot or cache age that your application can use as evidence. Ask for publication and effective dates, then check them at the source and state the answer cutoff.
Keep other proofs separate
A content mark can show that software produced a terminal text answer. It does not identify the web pages used or prove factuality. Evidence and audit explains what a task trace proves and what evidence your application must keep.
Scope and limits states the review requirements for high-impact use.