In a move to give web managers more control over their data, Google has announced a new feature that enables web admins to block AI systems from scraping their site content for generative AI training. This update comes shortly after OpenAI made a similar announcement about allowing web admins to block their systems from crawling content.

The Need for Control over AI Scraping

Web publishers have long been concerned about protecting copyright and preventing generative AI systems from replicating their work. With the discussion around AI regulation gaining momentum, it is becoming clear that stricter enforcement of data access for building generative AI models is on the horizon.

Google’s Solution: Google-Extended

Google’s new control feature, called Google-Extended, allows website administrators to choose whether to help AI models become more accurate and capable over time by controlling access to content on their site. This control extends to Google’s Bard and Vertex AI generative APIs, as well as future generations of models powering these products.

OpenAI’s Approach

OpenAI, too, has emphasized the importance of allowing web admins to block data access. OpenAI’s documentation states that retrieved content is only used to teach their models how to respond to user requests, not to make the models better at creating responses. This approach aligns with Google’s goal of improving AI models without compromising the integrity of web content.

Impact on Language Models

Although some large language models (LLMs) have already been built using data pulled from the web, web admins now have the opportunity to restrict access to their content moving forward. This development may limit the number of websites available for LLMs to access, which could have implications for search engine optimization (SEO) as more people turn to generative AI for web searches.

Google’s Experimentation with Generative AI

Google has been actively experimenting with generative AI in search through its Search Labs project. This initiative aims to explore the potential of generative AI in improving search results. As generative AI becomes more prevalent, websites may find it beneficial to be included in the datasets used by these AI tools to ensure their visibility in relevant queries.

The Importance of Web Admin Control

Google recognizes the increasing complexity web publishers face in managing different AI uses at scale. The company is committed to engaging with the web and AI communities to find the best way forward. By giving web admins more control over their data, Google aims to promote better outcomes for both parties involved.

Blocking Google’s AI Systems

Web admins who wish to block Google’s AI systems from crawling their sites can follow the guidelines provided by Google’s developers. These guidelines offer step-by-step instructions on how to prevent Google’s AI systems from accessing and scraping website content.

Future Considerations

As the field of AI continues to evolve, it is crucial for web publishers to stay informed and adapt to new developments. The ability to control AI scraping and access to content will likely become more significant as regulations and best practices are established. Web admins should be prepared to navigate the changing landscape of AI development and usage.


Google’s introduction of Google-Extended, a feature that allows web admins to block AI systems from scraping their sites for AI training, reflects the growing concern for data control and copyright protection. This move aligns with OpenAI’s efforts to give web admins more control over data access. As AI applications expand, web publishers will face the challenge of managing different AI uses at scale. Google’s commitment to engaging with the web and AI communities ensures ongoing collaboration and the potential for mutually beneficial outcomes. Web admins who wish to block Google’s AI systems can follow the guidelines provided by Google to protect their content and maintain control over their data in the age of generative AI.