HomeBlog › Claude vs ChatGPT Watermarking Compared

Claude vs ChatGPT Watermarking Compared

Editorial Team · published 11 August 2026

Quick answer

Anthropic ships a statistical text watermark on Claude models from 2 August 2026. OpenAI researched the same approach for GPT but has not deployed a comparable public text watermark. Google's SynthID marks images, audio and video. So "AI watermark" means a different mechanism depending on which vendor you mean.

“Does AI watermark its output” has no single answer, because the three major vendors made different choices and shipped them at different times.

The comparison

Anthropic (Claude)OpenAI (ChatGPT)Google (Gemini)
Text watermarkStatistical, token-level, live from 2 Aug 2026Researched, not deployed as a public checkable text watermarkNot a comparable public text watermark
Image / media markingC2PA manifests on file outputC2PA metadata on image outputSynthID, embedded in the pixels of images, audio, video
DriverEU AI Act Article 50(2) transparency codeVoluntary commitments, C2PA membershipC2PA membership plus in-house research
Public detectorForthcoming, not availableNot applicable for textLimited access
Survives copy-pasteYes, for textNot applicableYes, for media
Trivially strippedNo, for textMetadata yesHarder, embedded in pixels

Why the mechanisms differ

Statistical text watermarking biases which token the model picks among near-equivalent candidates, keyed to a secret. The signal is spread across a passage, so it needs length to be detectable and it degrades under heavy rewriting. It is the only approach that survives copy-paste out of a chat window, which is why regulation pushed toward it.

File metadata such as C2PA attaches a signed manifest to the file. Cryptographically strong as a record, structurally fragile as a marker: converting the format, re-saving, or screenshotting usually drops it. It is a provenance record, not a lock.

Pixel-embedded marking such as SynthID hides the signal in the image data itself, so it survives resizing, cropping and compression far better than metadata. It has no text equivalent that is publicly deployed.

What each one can actually prove

Less than the marketing implies, in every case.

  • A text mark shows the model touched the text. Not that it composed it. Claude may have proofread or summarised something a person wrote.
  • No mark proves nothing. The model may predate the rollout, the passage may be below threshold, or a different vendor was used.
  • Metadata absence proves nothing at all. One re-save removes it, and people re-save files constantly for entirely innocent reasons.

Anyone building policy on top of detection, in education, publishing or HR, needs those three limits written into the policy.

What this means if you are cleaning text

If your source was Claude, invisible-character cleaning does not touch the watermark. If your source was ChatGPT, there is no deployed public text watermark to remove in the first place, and what detectors flag is style rather than a mark. Either way, the tools selling zero-width deletion as watermark removal are mislabelling ordinary text hygiene.

The useful framing: clean characters because they break systems, and treat provenance as a disclosure question rather than a technical one. How to remove the Claude watermark, honestly.

Who this is not for

Anyone hoping one of these vendors left an easy exit. None of them did, and the differences between them are about what each mechanism can survive, not about which is easiest to defeat.

More detail on the Anthropic side: what Claude’s text watermark actually is.

Keep reading