# Google, Microsoft and OpenAI Agree to Honor One Shared robots.txt Opt-Out for AI Training

*2026-11-02 23:01 UTC · Technology · 1 min read*

> The three companies plan to honor a single robots.txt token, __usage_ai-training, though none supports it yet and rollouts run into 2027.

Google, Microsoft and OpenAI said they will honor one shared robots.txt signal, __usage_ai-training, that tells AI developers whether a site's content may be used to train generative models. The underlying Internet-Draft was written by Anthropic's Kevin Kelley for the IETF's AI Preferences working group. It is an individual draft, not an official standard, and expires on May 3, 2027 unless replaced.

None of the companies supports the signal yet. Google says it will treat the shared line as a Google-Extended block, with support targeted for early 2027. Microsoft will honor the new options from February 1, 2027, adds a page-level AI grounding preference, and will retire the NoArchive and NoCache tags on March 1, 2027. OpenAI aims to honor the signal within the next few months, while Anthropic says full support will take time.

The signal has limits: rules already in robots.txt files do not count, a wildcard group cannot match the token, and it does not control crawling. Critics such as Plagiarism Today's Jonathan Bailey note that it gives rights holders no new enforcement tools and does not affect material already collected.

## Sources

- [searchenginejournal.com](https://www.searchenginejournal.com/google-microsoft-openai-shared-ai-training-opt-out/595387/)
- [plagiarismtoday.com](https://www.plagiarismtoday.com/2026/11/02/ai-companies-announce-new-blocking-standard/)

---
TechTools News · https://techtools.cz/tools/news/article/2625-google-microsoft-and-openai-agree-to-honor-one-shared-robots-txt-opt-out-for-ai
Česká verze: https://techtools.cz/tools/news/clanek/2699-google-microsoft-a-openai-budou-respektovat-jednotny-zakaz-trenovani-ai-v