AI Disclaimer
Our credibility results are generated by artificial intelligence. They are a helpful starting point, not a final verdict. Please read this before relying on a result.
Results can be wrong
The analysis is produced by an AI model. Like all such models, it can be confidently mistaken. It may misjudge a claim, miss context, or produce a result that a human expert would disagree with.
Known limitations
- Satire and humour can be misread as sincere claims.
- Breaking news may be assessed before reliable reporting exists.
- Nuance and opinion are harder to score than plain factual claims.
- Bias present in training data or sources can be reflected in results.
- Source quality varies. A confident-sounding source is not always a reliable one.
Always verify important claims independently. Do not rely on our score alone for decisions about health, finance, legal matters, safety, or anything else that matters. Consult primary sources and qualified professionals.
How to use results well
- Read the explanation, not just the number. The reasoning is the useful part.
- Use the per-claim deep check to see cited sources for anything you want to probe.
- Treat a low score as a prompt to investigate, and a high score as encouraging but not conclusive.
Tell us when we get it wrong
Feedback makes the tool better. If you believe a result is mistaken, please challenge it and we will review it. You can read more about how scores are produced on our Methodology page.