Our position
AI tools are useful for drafting, summarising, cross-referencing and pattern-checking large volumes of publicly available information. They are not adequate for editorial judgement, regulatory interpretation, or unverified factual claims about a regulated financial-services provider. This policy sets out where the line falls.
Where AI is used
- Drafting first cuts of explanatory content that a human editor then reviews, rewrites and signs off.
- Summarising long primary documents such as regulator disclosures, provider terms or annual reports for editorial review.
- Consistency-checking scoring rationales across a large index of providers.
- Internal engineering, quality assurance and code review.
Where AI is not used
- Publishing any factual claim about a regulated provider without human editorial verification.
- Determining, adjusting or rationalising a provider's ranking score.
- Handling user data or generating responses on behalf of a named human editor without that editor's review.
- Producing statistics, figures or historical facts that cannot be traced to a primary source.
Mandatory human review
Every article, provider profile and policy on this site is reviewed by a named human editor before publication. The editor is responsible for the accuracy of the content and named on every article's byline. AI does not sign editorial content. If a piece of content changes materially, it is re-reviewed by the editor before the update goes live.
Transparency
We do not disguise AI-assisted content as unassisted, nor do we manufacture human bylines to hide AI involvement. The named editor is a real person whose review of the final content is a substantive act of editorial judgement, not a rubber stamp.