OpenAI will start watermarking ChatGPT’s text in the EU

OpenAI will begin adding an invisible watermark to text generated by ChatGPT and Codex in the European Union to comply with the EU AI Act, the company said Monday in a blog post.

The EU AI Act's transparency rules, which took effect on August 2, require AI companies to mark AI-generated content in a way other systems can identify.

OpenAI said the watermark will roll out over the coming weeks to eligible ChatGPT and Codex users on all plans, but only in the EU. Developers using OpenAI's API anywhere in the world can turn it on for select models starting today; it is off by default. OpenAI said it is not making text watermarking a global default at launch.

The watermark is not a symbol, but works by subtly shaping the model's word choices, leaving a pattern readers cannot see but a detector can pick up. Because it lives in the words themselves, it travels with the text when copied and pasted. OpenAI said the watermark does not identify the user, and that it saw no meaningful change in its models' performance with it switched on.

OpenAI also published a technical report for its method, called textGrain, co-written with researchers from the University of Pennsylvania and Yale. It describes using a secret key to sort next-word predictions to finish a sentence; adding hundreds of these nudges lets a detector spot AI-generated content using only the text and the key.

OpenAI's tests suggest editing can remove the watermark. In one test, replacing 10% of words with synonyms dropped detection from about 92% to 66%. The company also said short passages, math answers, and translated text are harder to detect. It said these limitations contributed to its decision to provide initial detector access only to approved researchers and expert organizations.

OpenAI cautioned that a missing watermark does not prove human authorship, and said watermarks can indicate that an OpenAI system generated or processed part of a passage, but not how much human judgment, editing, or creativity went into it.

Anthropic said two months ago that it would watermark text generated by Claude worldwide. The Wall Street Journal reported in 2024 that OpenAI had built a text watermark before but held off releasing it, partly over concerns users would switch to rivals that did not watermark.