Unverified50% confidenceEventExact time
Starting in October 2026, OpenAI will add invisible watermarks to the text output of ChatGPT and Codex in the EU, enabled by default for EU users and disabled by default in the rest of the world.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
VerifiedOpenAI is planning to add watermarks to text generated by ChatGPT in order to meet EU regulatory requirements for AI content provenance.71% similarUnverifiedOpenAI launched a text watermarking system called TextScreen, which embeds statistical signals in the model's word choices rather than inserting visible markers for subsequent detection, in response to the EU AI Act, and plans to open-source it.65% similarUnverifiedOpenAI is taking a regionally differentiated approach: watermarking is mandatory in the EU region and optional via API in other regions.64% similarUnverifiedStarting today, customers using the OpenAI API can manually enable text watermarking for certain models globally.64% similarUnverified闭源API服务（如OpenAI、Anthropic）可以在服务端强制嵌入水印而用户无法干预63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1042327API
curl https://kongchang.com/api/v1/knowledge/claims/1042327MCP
get_claim(id=1042327)