
What happened
OpenAI is adding the textGrain watermark to text from ChatGPT and Codex, applying it to EU outputs over the coming weeks. From October 5, API users worldwide can switch it on for some models, off by default.
Why it matters
The move responds to the EU AI Act's requirement that providers make AI-generated text machine-detectable, so outputs may carry a signal firms and regulators can check — though OpenAI says detection has limits.
What to watch
OpenAI will initially give the detector only to approved researchers and institutions, not the public, so real-world usefulness hinges on who gets access. Its own tests found detection drops to 66% when 10% of words are swapped.
WHO IT HITSCompliance, trust-and-safety, and platform teams at companies shipping generative AI products into the EU will need to account for a watermark that ships by default on EU outputs and is optional in the API. Editing tools and detection-reliant workflows may be affected, since OpenAI's tests show light rewording hurts detection.
Summaries like this, every morning.
The rollout is tied directly to the EU AI Act, which requires generative AI providers to make AI-generated text machine-identifiable. OpenAI is answering that with textGrain, a technique that avoids stamping characters or metadata onto the text and instead bakes a statistical signal into how the model picks its words.
OpenAI's own numbers show where the approach holds and where it frays. Detection is strongest on prose with wide word choice, but weakens when the text is short, when the subject offers few alternatives, or when a person edits it — swapping 10% of words in a 400-token English passage cut detection from about 92% to 66%, and swapping 25% dropped it to 17%. The company also stresses that a detected watermark cannot establish human involvement, ownership, responsibility, or user identity, and that a missing watermark does not prove human authorship.
That is why the detector will go first to approved researchers and institutions rather than the general public. The practical stakes look likely to hinge on how much access those outsiders get, and whether detection holds up outside OpenAI's tests — especially on text that has passed through editing tools.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Gomumari will start offering an AI agent on November 21, 2026 that reads English, Chinese and other inquiries…

Gomumari will start offering its 営業データ分析AIエージェント on 2026年11月21日

Gomumari will launch its Data Matching AI Agent on November 21, 2026; it reads ledgers and documents such as o…

MetAI published the five-block structure of its AI育成プログラム for companies, schools and local governments, coveri…

On OpenAI's MedChemBench, 99 problems drawn from 117 papers and patents covering 761 molecular structures, GPT…

AWS added a distributed KV caching skill to its Agent Toolkit for AWS, letting coding agents like Kiro and Cla…
