OpenAI will honor a generic robots.txt signal expressing AI training preferences
OpenAI、AI学習に関する共通のrobots.txtシグナルを尊重へ
On October 30, 2026, OpenAI announced that it will honor the general signal in robots.txt, allowing website operators to express AI training preferences to multiple participating AI providers at once. The signal is defined by the individual Internet-Draft draft-kelley-aipref-training-preference-signal-00, drafted by Anthropic's Kevin Kelley; its core is the cross-vendor policy token __usage_ai-training, it does not change the syntax of RFC 9309, and it has not yet received formal IETF status. OpenAI said the move is based on progress in the IETF AI Preferences working group, plans to implement the signal within the coming months and publish implementation details, and calls on other AI providers to follow suit; the existing independent controls for each crawler, such as GPTBot and OAI-SearchBot, remain in effect.