Short answer: ChatGPT text carries no deployed watermark. ChatGPT images do — every image generated since May 2026 includes C2PA metadata plus Google's SynthID watermark. The C2PA layer is verifiable right here.
Does ChatGPT watermark text?
No — as of the last verification date above, ChatGPT text output has no deployed watermark, on the free tier, Plus or the API. OpenAI researched cryptographic text watermarking and reportedly prototyped it, but never shipped it. Claims that a “ChatGPT text watermark detector” can identify ChatGPT output are not supported by anything OpenAI has deployed; AI-text classifiers are statistical guessers, which is a different thing entirely — see watermark checker vs AI detector.
What ChatGPT text sometimes does contain is ordinary hidden Unicode — special spaces or invisible characters introduced by the interface or by copy-paste. Those are inspectable:
Inspect ChatGPT text for hidden characters
Show the hidden characters inside the text
Copy cleaned text
Removes only the characters you select below (unusual spaces are replaced with a normal space). This is a transparency tool — it does not remove statistical watermarks and is not a way to hide AI use.
Official provider verificationClaude detection: announced
ChatGPT / OpenAINo text watermark deployed.
This page runs each official check the moment it becomes available — follow the rollout on the provider status page.
ChatGPT images: C2PA + SynthID
Since May 2026, OpenAI is a C2PA conforming generator: images from ChatGPT, the API and Codex carry signed C2PA Content Credentials and an invisible SynthID watermark; supported audio carries SynthID too. OpenAI runs its own public Verify preview tool, and the C2PA layer can be validated by any conforming tool — including this one, locally:
Check a ChatGPT image for Content Credentials
Keep in mind
Social platforms, screenshots and edits routinely strip C2PA metadata — a missing credential doesn’t mean the image isn’t from ChatGPT.
The SynthID layer inside OpenAI images can be checked through Google’s channels or OpenAI’s Verify tool, not locally.
Provider behavior changes fast; this page states its last verified date above and is updated from primary sources only.