How to Optimize News Content for AI Answer Engines

News is uniquely exposed to AI answer engines because time-sensitive queries are exactly what those engines answer with retrieval, and they now resolve many of them without sending a click. Pew Research (July 2025) found that when a Google AI Overview appears, users click a traditional search result just 8% of the time, versus 15% when no summary appears, and only about 1% of AI-summary searches produce a click on a source link inside the summary.

That shift is already hitting publisher traffic. Digital Content Next reported median year-over-year Google referral traffic down roughly 10% in mid-2025, with news brands down about 7%. The practical takeaway: being the source cited inside the AI answer is now the visibility that matters, so news AEO is about earning the citation, not just the ranking.

Freshness is effectively a retrieval filter for news, so accurate timestamps are non-negotiable. Stamp both datePublished and dateModified in ISO 8601 format with a timezone, and update dateModified whenever you make a substantive edit, because Google uses it as a documented freshness signal in Discover and AI Overviews.

All four major engines, Google AI Overviews, ChatGPT search, Perplexity, and Claude, rank sources with retrieval-augmented generation that weights recency, and multiple 2026 analyses report that a large share of AI-cited pages are under three months old. Refresh evergreen explainers instead of letting them decay, and make the visible publish and update dates on the page match your structured data.

Complete NewsArticle or Article JSON-LD is what lets AI engines attribute your reporting to you. Google's AI Overviews extract the author, headline, datePublished, and publisher directly from this markup for source attribution, so missing or malformed fields mean a lost citation or a mis-attribution to an aggregator.

Fill in every required property and keep the schema consistent with what is visible on the page.

Author and publisher authority is a trust signal engines weigh before citing reporting. Link each article's schema to a Person entity for the author and an Organization entity for the publisher, and give every author a visible byline, a credentialed bio page, and a stable profile URL that the entity references.

This reduces mis-attribution and strengthens E-E-A-T, and it matters because engines say they rely on editorial quality signals, not just popularity. A clear, consistent author identity across your site helps AI systems recognize your journalists as reliable sources.

Write so an engine can lift a self-contained passage that answers the query in the first sentence. AI systems preferentially cite content with hard data, statistics, direct quotes, and links to primary sources, because verifiable specifics add credibility to a generated answer.

Lead each story and section with the key fact, attribute figures clearly, and include original reporting, such as interviews, first-hand data, and named sources, that aggregators cannot replicate. Original, data-rich reporting is both harder to substitute and more likely to be cited.

Licensing deals are reshaping the news-AI landscape, but they are mostly available to large publishers today. Perplexity launched its Publisher Program in July 2024 with revenue sharing, and in July 2025 Gannett's USA TODAY Network signed a licensing agreement covering 200+ local markets and publications.

If you are not a major publisher, focus on the levers you control: complete schema, accurate freshness signals, author E-E-A-T, and original reporting. These are what make your journalism citable without a contract, and they compound as engines increasingly weight quality over raw popularity.