Synthetic Media Provenance and Consent Checklist

For the method in Synthetic Media: Check Provenance and Consent Before You Post. Two separate checks, both required, before posting AI-generated or AI-edited media of a real person.

Not for

Fully synthetic people, generic illustrations, or your own likeness with your own consent. If a real, identifiable person other than you could be recognized, use this checklist.

Part 1: Consent check (about the person depicted)

• The person knows this specific content exists
• They agreed to this specific use (not just the original photo/video)
• The context (private vs. public, personal vs. work) matches what they agreed to
• Future-use and takedown terms are documented, including channels you control and copies you cannot promise to recall

If any box is unchecked: message them before you generate or post anything.

Part 2: Disclosure check (about the audience)

• The applicable law, contract, venue, and platform disclosure rules were checked
• A reasonable viewer would not mistake this for an unedited photo or recording, OR it is labeled
• The platform’s own AI-content label is applied, where available
• A human-readable disclosure exists near the content, not only in metadata
• The disclosure names what actually changed (not a vague “enhanced”)

Provenance record (keep even if informal)

A valid Content Credential can verify integrity, binding, and signer/trust information for recorded assertions. It does not prove that every assertion is true, that the depicted person consented, or that the audience saw the disclosure.

Before you post

• Consent check complete (Part 1)
• Disclosure check complete (Part 2)
• Provenance record saved somewhere you can find it later
• If this is a deepfake made of you without consent, see the adult first-response plan instead