How we use AI

Archived article texts are kept as they appeared historically, cleaned of navigation and ads only. The words, the order, and the examples stay the same. No machine has rewritten them.

AI is used to draft new evergreen guides and site structure. A human editor reviews every draft, checks every claim against statute or circular, and rewrites anything that sounds like a chatbot or a Wikipedia summary.

We do not use AI to invent quotes, statistics, case studies, or expert opinions. If a guide includes a number, that number came from a named source or the public record. If it includes an example, that example is either hypothetical and labeled as such, or drawn from real council minutes.

The AI drafting process starts with a brief: the topic, the reader's question, the relevant law, and the practical detail that matters. The machine returns a structured draft. The editor cuts the puff, fixes the tone, and adds the concrete detail the machine cannot know.

We do not use AI to generate article ideas, headlines, or meta descriptions without human rewriting. The machine's first idea is usually bland, vague, or a clone of something already published. The editor's job is to make it specific.

Old articles are never fed into a machine for 'updating' or 'optimizing'. That would erase the historical voice and the context that made the piece useful in the first place. If a topic needs a fresh guide, we write a fresh guide.

AI does not choose what gets published, what gets corrected, or what gets taken down. Those are editorial decisions made by people who answer to the Association's members.

This policy will be updated if our tools change or if new risks appear. The principle will not change: machines draft, humans decide, and no text goes live without a named person taking responsibility for every sentence.