AI text watermarks: only Claude and Gemini mark what they write
Since this summer, text from Claude and Gemini carries an invisible mark. Text from ChatGPT, Grok, Mistral, Meta AI and models run at home carries none, and Chinese services settle for a visible label. Yet outside the companies involved, nobody can check a text today.
Can a teacher tell that an essay came from an AI? Can a recruiter tell that a cover letter was written by ChatGPT? Since this summer, the answer depends on the model. Two companies now slip an invisible mark, a watermark, into every text, and their own tools can find it again. The others do not, or not yet.
This article takes stock as of 2 October 2026: who marks, which models, what the mark actually tells you, and what language changes. It was written by Claude Opus 5.5, a model made by Anthropic, one of the companies involved, so it carries that watermark itself.
Europe has required marking since 2 August 2026
Who marks AI-written text, and since when
Text output of each service, as of 2 October 2026.
- Invisible watermark in the text
- Visible “AI” label (app, exported files)
- Announced
View as a data table
| Service | Text marking |
|---|---|
| Google · Gemini app | since May 2024 |
| Google · Gemini API | since August 2026 |
| Anthropic · Claude | since August 2026 |
| Chinese apps, in China | since September 2025 |
| Mistral | announced for 2 Dec 2026 |
| OpenAI · GPT | goal announced, no date |
| Meta · Muse, Llama | no text watermark |
| xAI · Grok | no text watermark |
| Open models run locally | no text watermark |
The push comes from Article 50 of the AI Act, which requires providers of generative AI to make their output detectable by a machine, text included. An image can hide that information in its file; pasted text keeps nothing. So the words themselves have to be marked: the model picks its words along a statistical pattern that no reader notices but that software holding the right key can recognise.
The code of practice that comes with the law sets the bar at 200 tokens, a good hundred words; below that, a mark would be too weak to be reliable and is not required. Providers must also offer a free checking tool, which none has yet opened to the public.
Model by model
The table covers the models in our ranking.
| Models in the ranking | Mark in the text | Who can read it |
|---|---|---|
| Claude Opus 5.5, Sonnet 5.5, Fable 5.1 and earlier versions | Invisible watermark, worldwide, in the app, the API and Claude Code | Anthropic, plus approved organisations: regulators, media, researchers, schools and universities |
| Gemini 3.8 Flash | Invisible watermark (SynthID), in the app since 2024 and in the API | Google; its checking tool is still waitlisted |
| GPT-6.1 Sol, GPT-6 Sol, Luna and Astra, GPT-5.6 | None (only images and audio are marked) | — |
| Muse Spark 1.3 (Meta), Grok 4.7 (xAI) | None | — |
| Mistral Medium 3.5 | None, promised for 2 December | — |
| DeepSeek, Qwen, GLM, Kimi, MiniMax, MiMo, Seed, Hy4 | Visible “AI” label in the apps, in China | Anyone, on screen |
| Open models run locally (Gemma, Qwen 27B, Granite…) | None | — |
xAI stands apart: it is the only major Western provider that has not signed the European code. It is still bound by the law.
What the mark actually tells you
For whoever receives a text, the mark is of little use today. Neither Google nor Anthropic offers a public tool, so a teacher cannot check anything alone, unless their school is granted the access Anthropic offers.
The mark is also fragile. It survives light edits, but a full rewrite, a paraphrase by another model or a translation by another tool erases it. Nor does it say who wrote the text: Claude may have drafted the whole letter or merely fixed three sentences of a human draft, and the mark looks the same.
Finally, the detectors sold to schools and companies do not read these watermarks. They are classifiers that estimate a probability from style, with known failures: a 2023 study found they flagged about 61% of essays by non-native English speakers as AI-written. A detector score is not proof.
Chinese models: a label on screen, in China
China got there first by another route. Since 1 September 2025, its rules require chat and writing services to show an “AI” label in the interface and to write one into exported files. DeepSeek complied on day one; in September 2026 the regulator told Doubao, Yuanbao, Qwen and Wenxin to apply it more strictly.
For text, that label stays on the screen: a passage copied from the chat window leaves without it. No Chinese provider has announced a hidden watermark in the words, and nothing is planned for their APIs or their international versions, the ones used from Europe or the US.
Models run at home: no mark
The watermark is applied on the provider’s servers, at the moment the model picks each word. A model running on your own computer, like those on our Local AI page, therefore carries none, whoever made it. The only exception is deliberate: whoever installs the model can switch on the marking tool Google published, with their own key.
Does the language matter?
Not for the obligation: the law targets services offered in Europe whatever the language of the text, and Claude’s watermark applies everywhere. An English text from Claude or Gemini is marked like a French or German one.
Language matters for how the mark gets lost. A text written in another language and translated into English by a different tool loses its mark, while a translation made by Claude carries Claude’s. No provider has published detection rates by language. Detectors differ too: according to its documentation as quoted by third parties, Turnitin scores English, Spanish and Japanese, but not French or German, which matters for students who write in more than one language.
Sources
- AI Act, Article 50; European Commission, Article 50 FAQ and signatories of the code (24 September 2026); Regulation (EU) 2026/1744.
- Anthropic, “Claude text watermark” (14 August 2026) and help page.
- Google DeepMind, SynthID in Gemini (14 May 2024); SynthID-Text and its limits; Gemini API forum (19 August 2026).
- OpenAI, content provenance; City AM, 12 August 2026 (OpenAI, xAI); Mistral, Vibe governance; Meta, TextSeal.
- CAC, labelling measures (14 March 2025); DeepSeek announcement; Qinglang review (2 September 2026).
- He et al., “Can Watermarks Survive Translation?”, ACL 2024; Liang et al., detector bias, 2023.
Conclusion
As of 2 October 2026, text written with Claude or Gemini carries an invisible mark; text written with any other model does not. The mark proves little to whoever looks for it: only Google and Anthropic can read it, it vanishes once the text is translated or rewritten with another tool, and it shows that a model took part, not that it wrote everything. The detectors sold to schools do not read it: they guess. Two dates will change the picture: 2 December 2026 for services already available in Europe, 2 February 2027 for shared detection.
