OpenAI will strengthen the traceability and verification system for AI-generated content, supporting EU AI transparency guidelines

Odaily reported that OpenAI has announced its support for the EU Commission's "Code of Practice on Transparency of AI-Generated Content," aiming to promote the development of a trustworthy AI ecosystem and enhance the traceability and transparency of AI-generated content.

OpenAI stated that this code is an important step in implementing the European Union's "EU AI Act." This support builds upon OpenAI's years of research, product development, and ecosystem collaboration around content provenance.

According to the introduction, OpenAI has been adding C2PA metadata to its DALL·E 3 image generation tool since 2024 to identify the source of AI-generated content. Since then, OpenAI has continuously improved its content labeling and detection technologies and launched public verification tools to help users determine whether an image contains source signals related to OpenAI-generated content. OpenAI indicated that it currently employs a multi-layered content provenance approach, including:

Adding C2PA Content Credentials to images generated by ChatGPT, Codex, and the API;

Combining C2PA metadata with SynthID watermarks in AI-generated images to enhance source identification capabilities;

Launching public verification tools to help users detect whether supported images contain source signals from OpenAI generation;

Participating in the development of the Coalition for Content Provenance and Authenticity (C2PA) standards to promote unified content provenance specifications across the industry.

OpenAI pointed out that AI content provenance technology is still in its developmental stage, and metadata may be lost due to operations such as uploading, downloading, format conversion, or screenshots, and watermarks can also be affected. Therefore, the entire industry needs to work together to promote more reliable and interoperable verification mechanisms.

OpenAI stated that it will continue to enhance content transparency in its products, participate in the development of open standards, and cooperate with the EU AI Office, member states, and the global AI community to promote the establishment of a safer, more transparent, and more trustworthy AI ecosystem.