OpenAI will start watermarking ChatGPT’s text in the EU
TechCrunch 1 week ago 20 ● 6 sources
OpenAI will add an invisible watermark to ChatGPT and Codex text outputs in the EU to meet the EU AI Act’s transparency rules. The rollout starts over the coming weeks, and OpenAI will let API developers enable it for select models starting today with it off by default. Watermarked text will carry a detectable pattern through copy-paste, while detector reliability can drop if about 10% of words are replaced with synonyms.