OpenAI is joining the Steering Committee of C2PA

The Coalition for Content Provenance and Authenticity (C2PA) develops technical standards for certifying the source and history (or provenance) of media content. C2PA project was initially started by Adobe, Arm, Intel, Microsoft and Truepic.

Today, OpenAI announced that it is joining the Steering Committee of C2PA. OpenAI will be contributing to the development of the standard. It is important to note that OpenAI started adding C2PA metadata to all images created and edited by DALL·E 3 early this year. In the future, OpenAI will be integrating C2PA metadata for Sora. Like other metadata, people can still remove the metadata from AI-created content, but they cannot easily fake or alter this information.

In addition to joining C2PA committee, OpenAI is also developing new provenance methods to improve the integrity of digital content. For example, they are developing tamper-resistant watermarking, tools that use AI to accurately identify the likelihood of content that originated from generative models. OpenAI is opening applications for access to OpenAI’s image detection classifier to a group of testers from today. Applications for accessing this new tool can be submitted here.

Finally, OpenAI has included audio watermarking into Voice Engine, its custom voice model, which is currently in a limited research preview.

“While technical solutions like the above give us active tools for our defenses, effectively enabling content authenticity in practice will require collective action. For example, platforms, content creators, and intermediate handlers need to facilitate retaining metadata in order to enable transparency for the ultimate content consumer about the source of the content,” wrote OpenAI team while making the above announcements.