OpenAI will invisibly watermark ChatGPT and Codex text in the EU to meet the AI Act

OpenAI announced on 5 October 2026 in a blog post, as reported by TechCrunch, that it will add an invisible watermark to text generated by ChatGPT and Codex in the European Union. The move is meant to comply with the transparency rules of the EU AI Act, which took effect on 2 August and require AI-generated content to be marked in a machine-detectable way.

How it works

• The watermark is not a visible symbol. It subtly shapes the model’s word choices using a secret key, leaving a statistical pattern that a detector can find in the text alone. Because it lives in the words, it survives copy-and-paste.
• OpenAI published a technical report on the method, called textGrain, co-written with researchers from the University of Pennsylvania and Yale.
• According to OpenAI, the watermark does not identify the user, and it saw no meaningful change in model performance.

Rollout

• Over the coming weeks, for eligible ChatGPT and Codex users on all plans — in the EU only. It is not a global default.
• API developers anywhere can turn it on for selected models from today; it is off by default.
• Detector access will initially go only to approved researchers and expert organisations.

Limitations, by OpenAI’s own account

In one test, replacing 10% of words with synonyms cut detection from about 92% to 66%. Short passages, maths answers and translated text are harder to detect. OpenAI also cautioned that a missing watermark “does not prove human authorship”.

Context

Anthropic said two months ago that it would watermark Claude text worldwide, which drew criticism from some users. The Wall Street Journal reported in 2024 that OpenAI had built a text watermark earlier but held off, partly over fears users would switch to rivals.