Large technology companies commonly utilize internet content to train their neural networks. In response to this practice, Google has introduced a new feature known as Google-Extended. This tool empowers website administrators to prevent Google from utilizing content hosted on their websites for training its Bard and Vertex algorithms.
Google-Extended Control
Google explains, “Today we are announcing Google-Extended, a new control that web publishers can use to control whether their websites help improve the performance of the Bard and Vertex generative APIs […] Using Google-Extended to control access to content on site, an administrator can decide whether to help these AI models become more accurate and capable over time.”
This control operates as a token that can be added to a website’s robots.txt file, which instructs search algorithms on what content should not be indexed. Website owners already have the option to include the appropriate token to block OpenAI’s scanner, responsible for AI bot ChatGPT and other neural network-based algorithms.
However, Google presents a unique case due to its status as the largest search engine company, notes NIXsolutions. The advantage of this new token lies in its ability to restrict the use of site content for neural network training while preserving site indexing. This ensures that a site’s ranking on the search engine remains unaffected. Simultaneously, website owners can have peace of mind knowing that their content will not contribute to the training of AI algorithms.