OpenAI has begun embedding C2PA Content Credentials into all images generated with DALL·E 3, according to a company blog post. The move is intended to make it easier to identify AI-generated images and trace their origin.
What's Changing
Content Credentials is an open technical standard maintained by the Coalition for Content Provenance and Authenticity (C2PA), a group that includes Microsoft and Adobe among its steering committee members. The standard embeds tamper-evident metadata into an image file describing how and when it was created — functioning similarly to a digital "nutrition label."
According to OpenAI, all images generated through DALL·E 3 in ChatGPT and via the API will now carry this metadata by default.
A Public Verification Tool
OpenAI has also launched a web-based tool that lets anyone upload an image and check it for C2PA Content Credentials, which can confirm whether the image was produced by an OpenAI model. The company says this is intended to give journalists, researchers, and the public a straightforward way to check the provenance of images they encounter online.
Why It Matters
Embedding provenance metadata directly into generated images is part of a broader industry push, backed by companies including Microsoft and Adobe, to make AI-generated content easier to identify. Content Credentials are not foolproof — metadata can be stripped by some editing tools or platforms — but they give researchers and platforms a mechanism to flag or verify content when it is preserved.