OpenAI will watermark ChatGPT outputs by default—but only in the EU
OpenAI will watermark ChatGPT text by default in the EU under the AI Act, and leave it off elsewhere.
OpenAI said it will automatically watermark text generated by ChatGPT in the European Union, while offering the same feature turned off by default elsewhere. The company tied the EU default to the EU AI Act's requirement that AI-produced content be marked so another tool can detect it. Its proprietary method, textGrain, inserts statistical patterns in word choices that a keyed detector can identify without an obvious quality change. OpenAI will initially give detector access to a limited group of researchers and organizations, with an approval process to add others. The company noted that existing schemes such as SynthID and C2PA remain relatively easy to circumvent.
- ChatGPT text watermarking will be on by default only in the EU.
- Outside the EU the same feature will be available but off by default.
- The change is meant to meet EU AI Act AI-content marking rules.
- OpenAI's proprietary method, textGrain, embeds detectable patterns in word choice.
- Detector access starts with a limited set of researchers and organizations.
Full article222 words · extracted from arstechnica.com · click to collapse
OpenAI will begin automatically watermarking text generated with ChatGPT in the European Union, the company has announced. It will also offer the watermarking feature in other regions, but it will be off by default outside of the EU.
The move in Europe is driven by a need to comply with the EU AI Act, which took effect in August. It requires marking content produced by AI models in a way that another tool can detect. Unfortunately, there is still no completely effective and reliable way to do that. A few standards already exist, like SynthID and the C2PA project, but they are relatively easy to circumvent for anyone with basic know-how.
The same is likely true for OpenAI’s watermark. Its method is proprietary; the company calls it textGrain, and has published a technical paper explaining how it works. But in general, it works like other LLM watermarking tools we’ve seen in the past: It puts patterns in the word choices that are not clear to a human reader, and that don’t meaningfully change the general quality of the output, but that someone with a key can use a specialized detector to find. OpenAI says it will be giving access to the detector to a limited number of researchers and organizations, and providing a request-for-approval process for others to be added over time.
Text extracted automatically; images, tables and formatting may be missing. Original: https://arstechnica.com/ai/2026/10/openai-will-watermark-chatgpt-outputs-by-default-but-only-in-the-eu/