Transparency notice

AI-generated content disclosure

Content on UK CopeCheck is generated by artificial intelligence and reviewed for methodology, not fact-checked line-by-line by humans. Model outputs may contain errors.

What is generated by AI

Automated systems collect candidate political and media claims, extract the claim and speaker attribution, compare it with public evidence, and draft the displayed score, mode, rationale, summary, and review flags.

Model families

The active pipeline uses the MiniMax model family for extraction, scoring, validation, and escalation. Each stored result records the model used where available, and historical items may name earlier model families.

Method in brief

1. Candidate claims are gathered from public sources and matched to a named speaker or institution.
2. The claim is assessed against a published rubric and public evidence, including ONS, Bank of England, and other UK public datasets where relevant.
3. Automated validation checks attribution, internal consistency, confidence, and whether the claim is actually scoreable. Weak or ambiguous results are held for review.
4. Published results retain source links so readers can inspect the original claim and evidence.

Scores are analytical commentary, not objective facts, and should not replace the underlying sources.

Corrections

To report a misattribution, missing context, data problem, or scoring error, use the feedback control on this site or the CopeCheck contact route. Include the page URL and the specific correction requested.