Google, Microsoft, OpenAI and Anthropic back one AI opt-out token

What happened: According to SiliconANGLE, Google, Microsoft, Anthropic and OpenAI plan to support one shared website crawling standard. Site operators could block AI training scrapers with a single robots.txt entry instead of one per provider.

The details: PPC Land reports that the shared token is named __usage_ai-training. The companies have submitted a draft to the Internet Engineering Task Force (IETF), which maintains the Robots Exclusion Protocol. The draft is an individual Internet-Draft dated 30 October 2026, and the IETF datatracker says it is not endorsed by the IETF.

Who's affected: Microsoft will honour the preferences from 1 February 2027, replacing the NoArchive and NoCache tags it has used since 2023. It is also adding an AI Grounding setting for pages an AI model reads to answer a user's question. Google targets early 2027 and will read the token as a Google-Extended opt-out.

The rivals: OpenAI plans support in the coming months, with similar controls. Meta has expressed support for the idea but has not committed to implementing it.

The catch: The token is a voluntary preference, not an access control. It does not replace GPTBot or Google-Extended rules. It applies only going forward, so content already crawled stays in existing training sets.

Why it matters: The standard would let operators opt out of several AI firms' training with one robots.txt line. It would not apply to models that perform exclusively non-generative tasks.

What's next: The draft must still pass through the IETF AI Preferences working group. Anthropic's rollout timing remains unconfirmed.