OpenAI announced Monday in a blog post that it will begin adding an invisible watermark to text generated by ChatGPT and Codex within the European Union. This move is intended to ensure compliance with the EU AI Act regulations.
Effective August 2, the transparency rules of the EU AI Act mandate that AI companies mark AI-generated content in a manner detectable by other systems.
OpenAI stated that the watermark will be rolled out to eligible users across all ChatGPT and Codex plans over the coming weeks, but only within the EU. Meanwhile, developers using the OpenAI API globally can enable the feature for specific models starting today, though it remains off by default. The company noted that text watermarking will not be a global default at launch.
Unlike a visible symbol, this watermark operates by subtly shaping word choices within the model's output. This creates a pattern that readers cannot see but detectors can identify. As the watermark resides within the words themselves, it remains intact when the text is copied or pasted. OpenAI clarified that the watermark does not identify the user and reported no meaningful impact on the models' performance when the feature is active.
The company also published a technical report detailing its new method, named textGrain. Co-authored with researchers from the University of Pennsylvania and Yale, the report illustrates how a secret key can be used to sort next-word predictions. By combining hundreds of these adjustments, detectors can identify AI-generated content using only the text and the key.
OpenAI’s tests indicate that the watermark can potentially be removed through editing. For instance, replacing 10% of the words with synonyms reduced detection accuracy from approximately 92% to 66%. Additionally, the company noted that short passages, mathematical answers, and translated text present greater challenges for detection.
“These limitations contribute to our decision to provide initial detector access only to approved researchers and expert organizations, who can help us evaluate reliability and responsible uses,” the company stated.
OpenAI also cautioned that the absence of a watermark “does not prove human authorship.” The text might be too short, heavily edited, or generated by a competitor's AI.
“[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,” the company added.
This announcement follows two months after Anthropic revealed its plan to watermark text generated by Claude, a strategy it is applying worldwide. This decision drew backlash from some users who argued that they had provided all the instructions, context, and decisions, while Claude was merely a tool.
The Wall Street Journal reported in 2024 that OpenAI had previously developed a text watermark but delayed its release due to fears that users might switch to competitors who did not implement similar measures.
Anthropic, Google, Meta, Microsoft, and OpenAI are among the companies that have pledged to adhere to the EU’s code of practice regarding AI-generated content.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.