Technology & AI
OpenAI to watermark ChatGPT and Codex text for EU users
OpenAI will start watermarking ChatGPT and Codex text for users in the European Union to comply with the AI Act. An invisible, machine-readable watermark in text output is rolling out, though editing can make these marks harder to detect.
Automated check, not yet reviewed by an editor. This story was assembled from the coverage below and its claims were checked automatically against the sources shown. Automated checks never mark a claim Confirmed or Contradicted, and statuses can change as new reporting arrives. How automated checks work
Live
What we know
OpenAI will start watermarking ChatGPT’s text in the EU.
Multiple sources confirm that OpenAI will begin adding an invisible watermark to text generated by ChatGPT in the European Union.
- Supports
OpenAI will start adding an invisible watermark to text generated by ChatGPT and Codex in the European Union
OpenAI will start watermarking ChatGPT's text in the EU | TechCrunch - Supports
An invisible, machine-readable watermark in text output is rolling out to ChatGPT and Codex, but only for users in the European Union at first.
OpenAI is adding text watermarking in ChatGPT and Codex | The Verge - Supports
ChatGPT will start including an “invisible watermark” on text outputs to comply with European Union regulations, OpenAI announced on Monday, but it will only apply to users in the EU.
OpenAI to Add Watermarks on ChatGPT's Text Outputs, But Only in the EU | PCMag
OpenAI will watermark ChatGPT and Codex text in the EU to comply with the AI Act.
The sources explicitly state that OpenAI will watermark text from both ChatGPT and Codex in the EU, and that this action is taken to comply with the EU AI Act.
- Supports
OpenAI will start adding an invisible watermark to text generated by ChatGPT and Codex in the European Union to comply with the EU AI Act, the company said Monday in a blog post .
OpenAI will start watermarking ChatGPT's text in the EU | TechCrunch - Context
Like OpenAI, Anthropic made the move to meet the requirements of the EU’s AI Act ;
OpenAI is adding text watermarking in ChatGPT and Codex | The Verge - Supports
ChatGPT will start including an “invisible watermark” on text outputs to comply with European Union regulations, OpenAI announced on Monday, but it will only apply to users in the EU.
OpenAI to Add Watermarks on ChatGPT's Text Outputs, But Only in the EU | PCMag
Editing can make the invisible marks harder to detect.
OpenAI's tests indicate that editing, such as replacing words with synonyms, can significantly reduce the detectability of the watermark. One source also mentions that editing can 'weaken' the watermark signal.
- Supports
Can the watermark be removed by editing? OpenAI’s tests suggest yes. In one test, replacing 10% of words with synonyms dropped detection from about 92% to 66%.
OpenAI will start watermarking ChatGPT's text in the EU | TechCrunch - Supports
Editing can also “weaken” the watermark signal.
OpenAI to Add Watermarks on ChatGPT's Text Outputs, But Only in the EU | PCMag
An invisible, machine-readable watermark in text output is rolling out to ChatGPT and Codex, but only for users in the European Union at first.
The sources confirm that an invisible, machine-readable watermark is being rolled out to ChatGPT and Codex, and that this initial rollout is specifically for users in the European Union.
- Supports
OpenAI will start adding an invisible watermark to text generated by ChatGPT and Codex in the European Union
OpenAI will start watermarking ChatGPT's text in the EU | TechCrunch - Supports
An invisible, machine-readable watermark in text output is rolling out to ChatGPT and Codex, but only for users in the European Union at first.
OpenAI is adding text watermarking in ChatGPT and Codex | The Verge - Supports
ChatGPT will start including an “invisible watermark” on text outputs to comply with European Union regulations, OpenAI announced on Monday, but it will only apply to users in the EU.
OpenAI to Add Watermarks on ChatGPT's Text Outputs, But Only in the EU | PCMag
What is disputed or wrong
No claims here are disputed or contradicted.
Not yet established
Every claim here has been settled one way or another.
Coverage reviewed
Reporting read while checking this story. Many outlets repeating one report still count as one source.
- OpenAI will start watermarking ChatGPT’s text in the EUTechCrunch · Oct 5, 2026, 08:36 PM UTC
- OpenAI is adding text watermarking in ChatGPT and CodexThe Verge · Oct 5, 2026, 06:08 PM UTC
Every change to this story
Nothing here is edited silently. Status changes, corrections and updates are listed with the reason given.
- PublishedAutomated
Published automatically. Not yet reviewed by an editor.
- Automated checkAutomated
Automated check: 4 claim(s) assessed against 3 source(s).
- Story openedAutomated