AI watermark database

Which AI models watermark their text

2 of these 5 assistants mark the text they generate, and ChatGPT does so in part. None of the marks is made of characters a scanner can find. This page records what each provider actually does, what the evidence is, and when it was last checked.

Text watermarking status by provider, with mechanism and verification date
Model	Text watermark	Mechanism	Last verified
ClaudeAnthropic	Marked	SynthID-Text, keyed token selection	2026-09-10
GeminiGoogle	Marked	SynthID-Text, tournament sampling	2026-09-10
ChatGPTOpenAI	Marked in part	textGrain, keyed word choice; EU and API opt-in	2026-10-11
CopilotMicrosoft	Not marked	C2PA on images, nothing for text	2026-09-10
GrokSpaceXAI	Not marked	Visible mark on media only	2026-09-10

What each provider does

Claude

MarkedAnthropic · verified 2026-09-10

Marked on models launched from August 2026, by a keyed bias in word choice rather than hidden characters.

Sources: Anthropic · Anthropic Help Center · TechCrunch

Read the Claude entry →

Gemini

MarkedGoogle · verified 2026-09-10

Marked in the Gemini app and web experience since 2024, by the most thoroughly published scheme of the five.

Sources: Nature 634(8035), 818–823 · Google DeepMind · Google AI for Developers · Google

Read the Gemini entry →

ChatGPT

Marked in partOpenAI · verified 2026-10-11

textGrain announced October 2026: opt-in in the API, rolling out to ChatGPT and Codex in the EU, no global default.

Sources: OpenAI · OpenAI, technical report · Rumi · OpenAI Developer Community · European Commission · European Commission

Read the ChatGPT entry →

Copilot

Not markedMicrosoft · verified 2026-09-10

No text watermark documented. Images always carry provenance metadata; audio and video marking is optional.

Sources: Microsoft Learn · Microsoft Learn · Microsoft Learn · C2PA

Read the Copilot entry →

Grok

Not markedSpaceXAI · verified 2026-09-10

Nothing documented for text. Images and video carry a visible label, and the maker did not sign the EU code.

Sources: SpaceXAI · SpaceXAI · European Commission · European Commission · City AM

Read the Grok entry →

Invisible characters are a different thing

The eight characters below are real, and they do turn up in text that has passed through a model, an editor, a PDF or a web page. They are not watermarks. Nobody places them to identify anything, removing them defeats no provenance system, and finding one points to a text-handling step rather than to an author.

These eight are what our tools detect and clean, and nothing more. The real text watermarks on this page work by biasing which word a model picks, which no scanner can read from text alone — so we cannot detect them, and neither can anyone else. What a rewrite does is replace those word choices, which is the only edit that reaches a mark made of them. Each vendor says as much in the pages cited above. What nobody outside a keyholder can tell you is what their detector would then conclude.

U+200BZero-width spaceRenders nothing, but splits the word for search and counting.

U+200CZero-width non-joinerBreaks a cursive join or ligature. Needed in some scripts.

U+200DZero-width joinerJoins glyphs, and holds emoji sequences together.

U+2060Word joinerPrevents a line break and leaves no visible trace.

U+FEFFByte-order markA file signature. Unicode discourages every other use.

U+00ADSoft hyphenRenders only at a line break, so it hides in plain sight.

U+00A0Non-breaking spaceLooks like a space, so a typed space will never match it.

U+2028Line separatorForces a line break. Legal, but rare in text a person typed.

Frequently asked questions

What is the difference between a watermark and an invisible character?

A watermark is a statistical signal placed while the model chooses its words; nothing is inserted, and reading it needs a key. An invisible character is a real code point that ended up in the text, usually by accident. Only the second can be found by scanning.

Can any tool tell me which model wrote a passage?

No public tool can, including ours. The providers that do watermark text keep detection behind a key, and the characters that scanners find are produced by editors and web pages as readily as by models.

Why do these pages carry a verification date?

Because the answers change. Anthropic began marking text in August 2026, and OpenAI announced a text watermark of its own on 5 October 2026, which turned one entry here from unmarked to marked. Each records when it was last checked against primary sources.

Invisible Character Detector

Unicode Cleaner

Working on your own draft?

GenPolish rewrites text you already have, showing every change before you accept it. A watermark lives in the word choices, so replacing the words is what reaches it. We show what share of your words changed, and make no claim about detectors.

Open GenPolish