Google: websites will be able to refuse to become an AI training center

Google has just formalized a function that will allow websites to refuse the absorption of their data for AI training purposes.

The issue of data swallowed in gigantic quantities by companies like OpenAI for their chatbot, without ever asking for authorization, has led to numerous controversies, some of which ended up ending in legal complaints. So to prevent human contributors to the internet from having their work used for free, and without their consent, in the future, Google has decided to offer a blocking function that should be all the rage!

Meet “Google-Extended”

Google is very committed to the question of artificial intelligence, with Bard obviously, but also with its next Gemini language model. However, the American firm seems to want to bring a few rules to the jungle of the web where companies shamelessly collect the data they need.

It has just formalized its new “Google-Extended” tool, which will allow content creators on the internet to let the search engine continue to index their site, while being able to prohibit the use of their data. by artificial intelligence developers.

It’s happening on robots.txt

How will this happen? Well, Google-Extender will be usable through robots.txt, the robot exclusion protocol placed at the root of websites which indicates which data can be indexed or not by crawling bots. Website creators will be able to indicate their desire not to have their data captured by the exploration bots of artificial intelligence companies.

This is a way of proceeding which had already been chosen by the New York Times this summer, when the American newspaper explicitly prohibited the indexing of its data by the OpenAI exploration bot in its robots.txt file. This important update from Google should thus achieve the same result, but without making indexing on the firm’s search engine impossible, an activity which is vital for most entities on the web.

Source : The Verge