OpenAI announced it will introduce an invisible watermark for text generated by ChatGPT and Codex in the European Union, aligning with the EU AI Act's transparency requirements. The watermark, which subtly alters word choices, will help identify AI-generated content without affecting model performance.
The rollout will begin in the coming weeks for eligible users in the EU, while developers globally can enable it for specific models. OpenAI emphasizes that the absence of a watermark does not confirm human authorship, highlighting the complexities of AI-generated text verification.