▸ Concept also: content authenticity, media provenance, AI watermarking

Content Provenance

The chain of evidence that lets you verify where a piece of content came from and whether it has been altered.

Learn first

In a nutshell

A piece of content has provenance when you can answer: who made it, when, with what tool, and whether it has been changed since. For AI-generated images, audio, and text this is newly hard — the output is indistinguishable from a human original, so the chain must be attached deliberately, either as a cryptographic manifest (C2PA) embedded at capture or as an imperceptible watermark baked in during generation. The hard part is that manifests can be stripped and watermarks can be attacked; neither survives all editing pipelines. Provenance is the infrastructure problem behind every AI-authenticity debate.

Where it came from

Year2021

SourceCoalition for Content Provenance and Authenticity (C2PA)

Why it matteredC2PA published its first specification in 2021, formalising the term for AI-era use. The underlying concept predates it in digital forensics and journalism.

In megatrends

Artificial Intelligence

Models, agents, and AI–human collaboration — general-purpose capability scaling into every domain.

Related players

◆ Nano Banana 2 ◆ SynthID

How this connects

Tap a node to open it