OpenAI has published details on how it plans to comply with European Union text provenance rules through watermarking technology. According to the company’s announcement, the approach focuses on where watermarks will be applied, how the detection system operates, and the initial rollout strategy for access to detection capabilities.
The company indicates that access to watermark detection tools will begin with researchers, though the announcement does not specify a broader timeline for public availability. OpenAI’s statement addresses the growing regulatory requirements in the EU around identifying AI-generated content, part of broader efforts to ensure transparency in artificial intelligence outputs.
While the announcement confirms OpenAI is developing a compliance framework for EU text provenance regulations, specific technical details about the watermarking methodology or which OpenAI products will include watermarks were not fully elaborated in the available source material. The researcher-first approach suggests the company is taking a phased implementation strategy as it works to meet regulatory requirements.