How should websites disclose AI-written advice without making readers play detective?

Pew Research Center found significant signs of AI authorship on 10% of a random sample of English-language webpages collected in July. Its methods paper makes the other point plainly: detector classifications of an individual page are probabilistic, not a definitive verdict about who wrote it. So readers should not have to inspect punctuation or trust a detector score to decide whether advice deserves attention. If AI helped write a page that could cost someone money, time or a bad decision, I want the page to say what a human actually checked: source links, publication date and who will correct it when it is wrong. “AI-assisted” on its own says almost nothing. What would make an AI-written advice page worth trusting to you?