AI Text Forensics

Hidden Unicode AI Watermark: Why Finding One Proves Almost Nothing

Invisible characters turn up in AI chat output constantly, and almost every one of them is a formatting artifact that predates chatbots by years or decades. No major provider has documented a watermark built on zero-width spaces, joiners, byte-order marks, or non-breaking spaces. Homoglyphs are the one category that deserves a second look, and even there the only concrete claim is a single unverified report about a discontinued Claude Code behavior.

Published 2026-08-11Updated 2026-08-11Community discussion

Key takeaways

Four characters explain nearly every find: zero-width space (U+200B), zero-width joiner and non-joiner (U+200D and U+200C), byte-order mark (U+FEFF), and non-breaking space (U+00A0). Each has a mundane, documented purpose unrelated to AI.
Zero-width joiners and non-joiners are load-bearing for Arabic and other complex scripts, and the joiner is what builds compound emoji. Byte-order marks are usually left behind by whatever software last saved or exported the text.
Homoglyphs are the genuine exception. A Cyrillic character standing in for a Latin lookalike has no everyday formatting purpose, and security researchers already treat character swapping as a real obfuscation technique.
The one concrete homoglyph claim: in a post dated July 2, 2026, developer Sean Goedecke argued text watermarks are "trivially removable" and wrote that Claude Code had once used Unicode homoglyphs in date strings as a since-discontinued steganographic flag. He said himself he wasn't certain how Anthropic's current system works, and the article he cited for corroboration is no longer reachable.
Community threads reacting to Claude's 2026 watermark confirmation list hidden Unicode as one guess among seven, none of which Anthropic's own support article names. Separate marketing claims about hidden Unicode in Grok's text have zero corroboration from xAI.
Finding one of these characters doesn't identify which model wrote the text, or that a model wrote it at all, unless a provider has published documentation of a scheme built on that exact character. None has.

Text watermark workflow

How text watermark removal is evaluated

Generate

Claude, Gemini, or another LLM produces text.

Mark

A watermarking layer can influence token choices or attach provenance.

Clean / transform

Text is normalized, edited, rewritten, or translated.

Check

A detector compares the before and after signal strength.

Why this matters for removal tools

Hidden-character cleanup is immediate. Statistical text watermark removal becomes stronger when paired with provider-specific detection and before/after confidence checks.

The four characters behind almost every find

Confirmed

Here's what is actually sitting in your clipboard, and why each character was there long before any chatbot existed.

Four characters account for the overwhelming majority of invisible-character finds in AI output.

Zero-width space (U+200B): no glyph, no width, used to create line-break opportunities inside long unbroken strings like URLs and hashtags.
Zero-width non-joiner (U+200C) and zero-width joiner (U+200D): control whether adjacent letterforms connect in Arabic and other complex scripts.
Byte-order mark (U+FEFF): written at the start of a file to signal byte order or confirm UTF-8 encoding.
Non-breaking space (U+00A0): keeps two words on the same line, like a name and a title.

Every one of those predates the modern chatbot, in some cases by decades. And each has an ordinary explanation for how it ended up in your text.

The zero-width space does have a less charitable use. Some content generators and spammers sprinkle it to pad word counts or slip past duplicate-content filters, which is a spam signal, not a provider signature.

The joiner is the one people underestimate. It's what glues emoji sequences together, so the family emoji and most skin-tone and profession combinations are literally built out of U+200D.

A byte-order mark is an encoding artifact left by whatever editor, export tool, or paste pipeline last handled the text. It tells you something about the software that saved the file and nothing about who wrote the words inside it.

The non-breaking space is the most common false alarm of the lot. Enormous numbers of web pages use it in their markup in place of a plain space, so any text copied from a browser is likely to carry a few.

Homoglyphs are the one category worth a second look

Community discussion

You'll see why a lookalike character is a genuinely different kind of find, and exactly how far the one concrete claim about it goes.

A homoglyph swaps a letter for a visually identical character from another script, like a Cyrillic character standing in for a Latin a. Your eye can't catch it. A Unicode-aware tool can.

Unlike a byte-order mark or a non-breaking space, there's no everyday formatting reason to do this. Security researchers already treat homoglyph substitution as a real technique, used in phishing domains and text obfuscation.

That's why one specific claim deserves naming. In a post dated July 2, 2026, developer Sean Goedecke argued that text watermarks are "trivially removable" and separately wrote that Claude Code had at some earlier point used Unicode homoglyphs inside date strings as a steganographic flag, since discontinued.

That's far more concrete than the usual folklore. It names a tool, a character class, and a use that has ended.

It's also unconfirmed. Goedecke said himself he wasn't certain how Anthropic's current implementation works, and the article he pointed to for corroboration is no longer reachable.

Why the leap from character to watermark doesn't hold

Community discussion

Here's the full list of things people guess Claude's watermark is, next to what providers have actually documented.

When Anthropic confirmed a Claude text watermark in 2026, the same follow-up question filled every comment section: how does that even work for text? The theories piled up fast.

hidden Unicode or invisible characters
statistical word-choice patterns
overrepresented n-grams
first-letter or sentence-position patterns
token-probability nudges
something SynthID-like
a hybrid of several signals at once

Anthropic's own support article names none of them. Hidden Unicode is one guess out of seven, and it happens to be the only one you can test for at home, which is probably why it gets checked first and reported loudest.

The same pattern shows up elsewhere. Marketing pages for the removal tools gptcleanup.com and scriptbyai.com have claimed hidden Unicode in Grok's text output, a claim with zero corroboration from xAI.

Compare that with getgpt.app, whose scanner checks for 34 or more specific invisible characters and states plainly that it does not detect or remove cryptographic watermarks like SynthID. Same underlying capability, honest label.

What to actually do with a character you found

Confirmed

Here's how to clean it up, and how to be clear-eyed about what the cleanup accomplished.

Stripping invisible characters is worth doing on its own merits. They break search-in-page, corrupt pastes into code and spreadsheets, and occasionally trip duplicate-content checks.

This site's cleaner runs entirely in your browser and reports per-character counts across eight of them: U+200B, U+200C, U+200D, U+2060, U+FEFF, U+00AD, U+00A0, and U+202F. Nothing is uploaded anywhere.

What it won't do is touch a statistical watermark, because there's no character to delete. Cleaning invisible Unicode and defeating a watermark are separate jobs, and only one of them is available to you in a text box.

FAQ

If I find a zero-width character in text I copied from an AI chatbot, does that prove which model wrote it?

No. Zero-width spaces, joiners, byte-order marks, and non-breaking spaces all have ordinary technical explanations with no connection to any specific AI provider. Unless that provider has published documentation of a watermark built on that exact character, its presence identifies nothing, and doesn't even confirm the text is AI-generated.

Are lookalike Unicode characters as harmless as zero-width spaces?

No, and it's worth separating the two clearly. Homoglyphs have no everyday formatting purpose, which is why security researchers treat character swapping as a real obfuscation technique. One unconfirmed report claims Claude Code briefly used homoglyphs this way in date strings, but that's a single blogger's account of a feature that no longer exists, not documentation of any watermark running today.

Does finding a hidden character at least confirm the text passed through some AI tool?

Not reliably. Word processors, CMS pipelines, and ordinary web copy-paste all insert these characters with no AI involved. A stray zero-width space or byte-order mark is a weak signal of almost nothing, let alone proof of which system produced the text.

Has any provider ever documented a Unicode-based watermark?

Not one. Google documents SynthID Text as a change to token sampling during generation, and Anthropic describes weaving an imperceptible watermark into the text itself. Neither describes inserting a character you could find and delete.

Next steps

Run a suspicious passage through the free in-browser cleaner and see exactly which characters are in it, with counts per character. Scan your text
Read how zero-width steganography actually encodes a hidden message, and why deleting the characters destroys it completely. Zero-width space watermarks
Check what Anthropic has and hasn't said about Claude's real, confirmed text watermark. The Anthropic watermark, sourced
If it was punctuation rather than an invisible character that made you suspicious, start with the em dash myth. Do em dashes mean AI?