Mechanism: announced/experimentalTWD integration: unsupported
OpenAI text watermark status
OpenAI developed a text watermarking method but is still evaluating it alongside metadata and other provenance options.
Last verified: 2026-08-17
Verification availability
What the provider publicly documents for text.
Known method
Research text watermarking method
Public verification
OpenAI does not document a public verifier for text. Its current verification tools cover supported image and audio signals.
Limitations
What to keep in mind when reading a result or claim.
- OpenAI reports reduced robustness under translation, extensive rewriting, and some deliberate text transformations.
- OpenAI has not documented the research method as a watermark on production ChatGPT or API text.
Authoritative sources
We use the provider's own documentation, announcements, and research. We do not use claims from third-party detector vendors.