Most publishers today use AI to write articles, create fake authors, and generate stock photography of people who don’t exist. We don’t — not because we can’t, but because our readers trust us to know the difference between a human who plays golf and a language model that predicts the next word.

What You Will Never Find On This Site

Where We Do Use AI

AI is a tool. Used poorly, it’s a shortcut to generic content. Used well, it’s a spellcheck that never tires. We use AI in a narrow, disclosed set of workflows:

1. Grammar and style checks. We run drafts through Grammarly and similar tools. The author accepts or rejects every suggestion — the tool doesn’t make the change.
2. Transcript cleanup. When we interview a tour pro or a fitter, we use AI transcription (Otter, Whisper) to produce a rough transcript. A human editor then cleans it, verifies against the audio, and ships the corrected version.
3. Alt-text drafting for accessibility. We use AI to draft image descriptions for screen readers, then a human editor reviews every one before publishing.
4. Internal research summaries. When an editor wants a quick recap of 10 years of USGA rule changes, an AI may produce a summary — which the editor then verifies against the primary rulebook before anything enters the CMS.
5. Internal code and spreadsheets. Our testing tools, scoring calculators, and data pipelines sometimes use AI assistants for engineering. No AI-assisted code touches what you read — only what we measure with.

What Changes If We Ever Use AI in Published Content

We can’t rule out the future. If we ever publish AI-assisted text — say, a real-time match report generated from live Tour data — the article will carry a visible, top-of-page disclosure that includes:

• Which model or tool was used.
• What inputs the human provided.
• Who the human reviewer was (named, credentialed).
• What role AI played (summarisation, drafting, translation, etc.).
• A link back to this policy.

We will not retrofit this disclosure to older articles, because there is nothing to disclose.

On AI and Search

Search engines increasingly use AI-generated answer boxes that may summarise our content. We cannot control that — but we design our articles to be clearly attributable, so that if our words appear in someone else’s summary, it’s obvious where they came from. We also publish structured data (author credentials, review dates, methodology links) so that AI systems and humans both have the full context.

If You Suspect AI-Generated Content

If you ever read an article on this site that feels like it was written by a machine, tell us. Some tells — excessive listicle structure, generic adjectives, no named experiences, stock phrases — can creep in from tired writers as easily as from language models. Either way, we want to know.

Contact the editor with the URL and what raised the flag. If a human editor agrees, we’ll rewrite or retire the piece and log it in our Corrections Policy.

The Promise

This policy is a promise, not a marketing slogan. If you ever have reason to doubt it, we’ll prove the provenance of any article on this site — original draft history, editor notes, author conversations, photography source files. We keep them all.