# The New Stack — OpenAI brings text watermarking to its API — and unlike Anthropic, it’s off by default

- Company: The New Stack (thenewstack.io)
- Announced: 2026-10-05T19:29:15+00:00
- Category: capability-change
- Coverage: not counted
- Announcement: yes
- Group: announcements
- Source: https://thenewstack.io/openai-api-text-watermarking/
- Record: https://forck.live/items/16090-openai-brings-text-watermarking-to-its-api-and-unlike-anthropic-it-s-off-by
- Subject: Reporting

OpenAI has introduced text watermarking for API-generated text, called textGrain, which embeds a statistical signal by subtly influencing word choices. The feature is opt-in and off by default for API customers globally, giving developers control over its use. OpenAI plans to automatically watermark eligible text from ChatGPT and Codex in the EU in response to the EU AI Act.

## Evidence

Verbatim from https://thenewstack.io/openai-api-text-watermarking/:

> "Starting today, API customers globally will be able to opt in to text watermarking for select models. Text watermarking will remain off by default in the API," the company writes. "This lets customers decide how watermarking fits their transparency obligations and the experiences they provide to users."

---

Record: https://forck.live/items/16090-openai-brings-text-watermarking-to-its-api-and-unlike-anthropic-it-s-off-by
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md