Google gives itself the right to use everything you post on the Internet to train its AI

Hoping to create a truly omniscient AI, Google changed the small letters of its privacy policy to grant itself the right to use absolutely all content available online to form its very large language model, Bard.

When, on its Privacy Policy and Terms of Service page, Google said it does its best to protect our information while allowing us to keep control of it. What he did not state in black and white was that his policy regarding scraping, data harvesting, has totally changed on July 1. In the race to offer the most powerful Artificial Intelligence between the online search giant and Microsoft, Google will use its dominant position to feed its AI with ever more recent public content.

To read – Windows 11: how to activate Copilot and take advantage of AI on your PC

Google does not hide it: it collects information about us to offer ever more personalized services and better quality. With its new confidentiality policy, it grants itself the right to use all the content, which it helps to index, published on the Internet. The new policy applicable on July 1, 2023 states: “we use publicly available information to help train Google’s AI models and building products and features like Google Translate, Bard, and cloud AI capabilities.”

Google gives itself the right to use all content on the web to train its AI

Note that Bard is not officially available in France for the moment. Many experts explain this absence by the fact that European laws are too restrictive for Google, which cannot apply them as part of its AI. If you live on the Old Continent, then, you have to go through a VPN to test the chatbot of Mountain View.

In addition, while waiting to be able to offer Bard in France and in the EU, Google is testing a new type of search engine giving results combining answers generated by the in-house conversational AI and organic results.