Editorial transparency
How hants.online uses artificial intelligence
We use AI as an editorial and research tool. It does not replace evidence, attribution or publisher responsibility.
Plain-language disclosure
AI is part of our workflow
hants.online uses generative artificial intelligence to help research public information, identify relevant official documents, summarise lengthy material, compare claims and assist with drafting.
We also use AI to neutralise loaded, promotional or partisan language. This means rewriting material in clearer and more even-handed terms while retaining justified scrutiny, uncertainty and attribution. Neutral language does not mean withholding criticism where the evidence supports it.
What AI may help us do
- find and organise government, Parliamentary, legislative and council publications;
- extract relevant passages, dates, figures and recurring themes for checking;
- compare proposals, consultation responses and official statements;
- draft, edit and simplify explanations;
- identify language that is emotive, promotional, ambiguous or insufficiently attributed;
- support website development, testing and accessibility review.
What AI is not permitted to decide
AI output is not treated as evidence or cited as a source. It does not determine our editorial position, decide whether an allegation is true, or remove the need to inspect the underlying publication. Content is not intentionally published solely because an AI system generated it.
The publisher remains responsible
Responsibility for selection, wording, publication and correction remains with hants.online. An error is our error even if an AI tool contributed to it.
Verification standard
Check the source, not the machine
Consequential factual claims should be checked against primary or authoritative material, with links provided so readers can inspect the evidence. We prefer legislation and Parliamentary records, followed by government publications, official council records, audited documents and recognised public datasets.
Where the evidence does not establish an answer, we should say so. Predictions, interpretations and accountability judgements are distinguished from confirmed facts using the site’s factual-report and analysis labels.
AI systems can invent details, omit context and reproduce bias. Source links, careful attribution and explicit uncertainty are therefore required safeguards, not optional additions.
Read the editorial methodologyPrivacy and disclosure
Readers should know what the tools touch
Visitor information
We do not intentionally submit newsletter addresses, postcode-checker entries or other visitor personal information to generative AI services. The public website does not contain an AI chatbot or send readers’ page activity to an AI provider.
Routine editorial assistance
Because AI is used as a routine research and editing tool, every AI-assisted sentence is not individually labelled. This policy provides the standing disclosure. Material AI-generated images or media will be identified where they are published.
Corrections
Readers may challenge a claim regardless of how it was produced. Material factual errors will be corrected under our editorial methodology; use of AI is never a reason to avoid or weaken a correction.
Current tool disclosure
The initial research, drafting and website development for hants.online were assisted by OpenAI’s Codex. This disclosure will be updated if a materially different generative AI system becomes part of the publishing workflow.
Policy review
This policy will be reviewed when the site adopts a materially different AI tool or use, or when its editorial or privacy implications change.
Published and last reviewed: 27 June 2026.