AI content policy

We use large language models in the editorial process. This page tells you exactly what for, what we never do, and how a page reaches you.

What we use AI for

• Outline drafting from research notes.
• First-draft copy that an editor then verifies and rewrites.
• Copy-editing for clarity, grammar and reading level.
• Generating comparison-table skeletons that an editor populates with verified data.
• Summarising vendor documentation as a research input — never as the published text.

What we never use AI for

• Generating user reviews, testimonials, ratings, or expert quotes.
• Fabricating screenshots, dashboard data, or pricing.
• Auto-publishing pages without a human editor pass.
• Producing pages in bulk for keyword variations without unique value (Google's spam policies forbid scaled content abuse regardless of method).

Editorial process per page

1. Research notes are compiled from primary sources (vendor docs, ad-platform docs, independent reviews).
2. An AI assistant drafts a structured outline from those notes.
3. A human editor writes or rewrites the body, adds judgement, removes hallucinations and inserts verified data.
4. A second pass verifies every numeric claim against a source.
5. Affiliate links are inserted with rel="sponsored noopener"; disclosure banner placed above the fold.
6. Published. Date stamped. Review-due date set to 90 days.

Why we tell you

Google's policy on AI-generated content is that quality matters, not method. The risk is not "AI was used"; the risk is "unhelpful content was produced at scale". We publish this policy so the standard is checkable. If you ever see a page that looks like a raw LLM dump, tell us at editor@clickcease.study and we will rewrite or unpublish it.