Google's New Token Allows Publishers to Control AI Content Use
The Google-Extended token is a new way for websites to control how their content is used by Google's AI models. It was introduced in September 2023 as part of a public consultation, and allows publishers to decide whether their sites help improve Bard and Vertex AI generative APIs.
According to the documentation, the token does not affect inclusion or ranking in Search, but it can be used to refuse one use without vanishing from the other. For example, a site can opt out of training future generations of Gemini models while still allowing its content to be used for grounding, which provides factuality and relevancy to chatbot answers.
However, there are some limitations and disputes surrounding the token. One issue is that it does not provide a clear separation between AI use and search visibility, and uptake has trailed behind rivals such as OpenAI's GPTBot.
The Independent Publishers Alliance, Movement for an Open Web and Foxglove have complained to the European Commission about Google-Extended, citing concerns over scope and bundling. Penske Media also sued Google in September 2025, alleging that the company trains search-specific products on content opted out through Google-Extended.