A verifiable record, attached to a piece of media, showing what created it, what tools touched it, and whether it has been altered since, so anyone can check whether an image, video, or audio file is what it claims to be.
Also known as Content Credentials, digital provenance, C2PA
Content provenance is metadata that travels with a file and answers a simple question: where did this come from? The Coalition for Content Provenance and Authenticity, or C2PA, is the open standard behind most of this work, cryptographically signing a record of which tool created a piece of content, what edits were made, and whether the file has been tampered with since it was signed.
By 2026 the coalition includes more than 6,000 members, among them Google, Microsoft, Adobe, Meta, OpenAI, Sony, and major news organizations, and the tools people already use, Adobe Firefly, OpenAI’s image generators, Microsoft Copilot, TikTok, Meta AI, emit Content Credentials by default. It sits alongside watermarking approaches like Google’s SynthID, which embed an invisible signal directly into pixels or audio rather than attaching a separate record.
Provenance used to be a publisher’s problem. Now it is a creative production problem too, because clients want proof of what is real, what is AI assisted, and what is fully generated, and platforms are starting to require that disclosure.
Disclosure is becoming mandatory, not optional. Meta, YouTube, and TikTok already require AI content labeling on political and social issue ads, and several states have their own rules. An agency that does not track provenance risks shipping creative that violates a platform policy it did not know existed.
It is a trust signal clients are starting to ask for. Some brands want Content Credentials attached to campaign assets as proof of authenticity, especially after a competitor gets caught passing off AI work as fully human made.
It changes the production checklist. Knowing which tools in your stack emit C2PA metadata, and which strip it out during editing or compression, is now part of a responsible AI assisted workflow, not an afterthought.
A political advocacy client asks your agency to produce a series of video ads using AI generated b-roll footage blended with real interview clips. Before the campaign ships, your production lead checks whether the final export still carries Content Credentials from the AI tools used, since several video platforms will reject or restrict ads that cannot verify their AI involvement once flagged. The team documents which shots are AI generated, which are stock, and which are original footage, then attaches that record to the client deliverable alongside the final cut. When a fact checking outlet later asks the client to verify a specific clip, the agency can produce the provenance trail in minutes instead of scrambling to reconstruct a production history from memory.
The automations and agents module of the workshop teaches you how to build AI workflows that compress the busywork without taking the craft out of the studio.