OpenAI, Google, and Microsoft to support single robots.txt line for AI opt-out
On October 30, OpenAI, Google, and Microsoft announced support for a new mechanism allowing website owners to refuse AI training on their content using a single line in the robots.txt file. The proposed standard, drafted by an Anthropic representative, uses the User-Agent string "usageai-training" to signal refusal to all participating companies at once, replacing the need to list specific bots…
Key points
- OpenAI, Google, and Microsoft announced support for a unified robots.txt AI training opt-out signal on October 30.
- No companies are reading the "usageai-training" signal yet; implementation starts in 2027 or "within the next few months".
- Using the opt-out line may remove content from Gemini answers unless Google-Extended is explicitly allowed separately.
However, the feature is not yet active. As of November 1, 2026, no companies are reading this signal. Microsoft plans to start on February 1, 2027, while Google targets early 2027 and OpenAI says it will implement it within the next few months. Anthropic has not set a date. The article notes that a generic "*" wildcard does not count as a refusal, and the string requires two leading underscores to be valid.
A key caveat for site owners is that using this line may also remove content from Google's Gemini answers, as it is treated similarly to rejecting Google-Extended. Those who want to block training but remain visible in AI search results must explicitly allow Google-Extended separately. The author, who manages a blog on a platform that controls its own robots.txt, highlights the importance of owning a custom domain to have this control.
The story so far
2 episodes →- OpenAI, Google, and Microsoft to support single robots.txt line for AI opt-outthis story
OpenAI and others announce support for a single line to 'opt out of AI training'
note.com · 4 November 2026
Loading the full article…
This text was published by note.com and written by kinu. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Policy & Regulation
All →- Google says schools must accept AI features to retain Search access for K-12 students · 18 src
- Democratic wins may expose 50% of US AI data center plans to scrutiny · 1 src
- Cert‑in reportedly gains access to Anthropic’s Mythos AI model · 1 src
- Democratic sweep of US Congress drags Asian tech stocks · 1 src
- Irregular Security blames human error for AI hacking incidents · 1 src
Comments
via GitHub Discussions