AI firms plan a shared publisher opt-out signal for training-data scraping

OpenAI, Google, Microsoft and Anthropic pledged to support a standard way for publishers to signal that their content should not be scraped for future AI model training. The rollout timelines differ, and the signal will not stop other crawling; scraping remains allowed by default until a publisher changes its setting.

• The proposed standard would let publishers use one instruction, rather than separate blocking commands for each bot, to signal that their content should not be used for future AI training.· 1 source
• The companies gave different implementation timelines: Microsoft said it would begin on February 1, 2027, Google sometime in 2027, OpenAI in the coming months, and Anthropic said rollout had started.· 1 source
• The setting applies to future model training, does not prevent other crawling, and leaves scraping allowed by default until publishers change the setting.· 1 source