Google Updates Gemini/Vertex AI User Agent Documentation

Google has updated documentation for a user agent expanded by Google, which publishers can use to control whether Google Gemini and Vertex uses their data for training or for the basis of AI response.

Updated guidelines

Google updated its Google Guidelines Extended based on the feedback of publishers to improve clarity and add more specific details.

Previous documentation:

“Google-producing is a solo product of products that web publishers can use to manage whether their websites improve twins and VERTEX AI generative APIs, including future generations of models that start these products. Taking Google Search on VERTEX AI does not use grounding websites rejected by Google-Opsed.

Updated version

The updated documentation provides more details and it is easier to understand the explanation of what the user agent is and what blocks it.

“Google-producing is a solo product of products that web publishers can use to manage whether Google Puzi’s contents from their web site training for future Bemini models that are powered by BeMina Apps and Vertex AI API for Gemini and the foundation (providing content from Google search in the work mode and relevant work) in Geminshins) in Geminshins) faction) in the mode of work in the Gemins model) in factual) in Gemins) in Gemins.

Google-Propagated is not a ranking signal

Google has also updated one sentence to clearly say that Google Exted is not used as a ranking signal for Google Search. This means that the allowing Google extended use of data to ground the twin response of AI will not count as a ranking signal.

Grounding is a reference to the use of web data (and a database of knowledge) to improve the response to a large linguistic model with up -to -date and factual information, helping to avoid fabrications (also known as hallucinations).

The previous version omitted mentioning the ranking signal:

“Google-Propagated does not affect the on or ranking of the website in Google Search.”

The new updated version specially mentions Google extended in the context of ranking signas:

“Google-EXTEDED does not affect the website on Google search or used as a ranking signal in Google search.”

Documentation corresponds with another leadership

The updated documentation coincides with a short passage of extended Google that is elsewhere in the Google Search Central. Another longtime guideline explains that Google Transuded is not a way of controlling the way the information about the web site is shown in Google search, showing that Google-Propagated is separated from Google search.

Here Other leadership This is located on the prevention page for the content to appear in Google AI reviews:

“Google-Propagated is not a method of managing the way your content appears in Google Search. Instead, use other methods to manage content in search, such as robots.txt or other robot controls.”

Taket

• Documentation update with extended Google:
  The documentation extended to Google is clarified and expanded to facilitate its purpose and effects.
• Separation from ranked signals:
  Updated guidelines explicitly state that Google prolonged does not affect Google searching, nor is it a ranking signal.
• Internal use of AI model:
  He explained that Google expanding if the content of the website for training and grounding the twin model is controlled.
• Consistency through documentation:
  The updated language now corresponds to the long -standing guidelines elsewhere in Google’s documentation, enhancing the separation from the search visibility control.

Google has updated its extended Google documentation to explain that publishers can block their content to be used for AI training and grounding without affecting Google’s search. The update also coincides with long -standing guidelines explaining that Google prolonged has no effect on how websites are indexed or ranked in search.

Sepaled picture Shutterstock/Jhvephoto