OpenAI C2PA Policy (help.openai.com)

OpenAI has announced its implementation of the C2PA standard and its own SynthID watermarking technology, enhancing the verification of media origins for AI-generated images produced through tools like ChatGPT and Codex. The C2PA standard allows for the embedding of metadata that certifies the source and provenance of various media, a move also embraced by camera manufacturers and news organizations. Meanwhile, SynthID provides an invisible watermark that persists even through common edits, ensuring a reliable provenance signal alongside the detailed context that C2PA metadata offers. This dual approach is significant for the AI/ML community as it addresses concerns over the authenticity and origin of AI-generated content. By enabling users to verify images through the presence of embedded metadata and watermarks, OpenAI enhances trust and transparency in AI-generated media. Users can confirm the origin of their images via a dedicated verification tool, although it does not guarantee the accuracy or legality of the content itself. This initiative highlights the importance of provenance in an age where AI-generated images are increasingly prevalent, fostering a more responsible use of AI technologies in media creation.