Everything On The Internet Can Be Used For Free To Train AI Models: Microsoft's AI Division CEO

Microsoft has joined AI companies like Perplexity, Google and OpenAI who think that it is okay to train their AI models on the content which is available on the web without having to pay the creator.

Numerous AI companies are now being surrounded by controversies about scraping copyrighted data and using it to train their upcoming AI models.

The Chief Executive Officer of Microsoft's new artificial intelligence (AI) division, in a recent interview with CNBC, said that anything that is published on the internet ends up becoming a 'freeware' which can be copied and then used for the training of AI models. He was asked if AI companies have effectively stolen the world's IP or not.

Mustafa Suleyman responded by saying, “With respect to content that is already on the open web, the social contract of that content since the 90s has been that it is fair use. Anyone can copy it, recreate with it, reproduce with it. That has been freeware, if you like. That’s been the understanding.”

Microsoft Sidelines With AI Companies Against Publishers & News Organisations

He then went on to say that all the content available online can be used freely to train AI models unless the publisher of that content or a news organisation explicitly asks Google to not scrape or crawl their content for anything other than indexing in order to make content visible to other people. This has clearly made Microsoft's stance clear on the matter of the use of content uploaded on the internet to train AI models.

By doing this, the tech giant has joined the side of AI companies like Perplexity, Google and OpenAI who believe that it is okay to train their AI models on the content which is available on the web without having to pay the creator.

Almost all companies working in the field of generative AI such as ChatGPT, Google Gemini, and Copilot are being surrounded by controversies about scraping copyrighted data and using it to train their upcoming AI models.

In addition to these controversies, numerous organisations and publications including Forbes, the New York Times and the Recording Industry Association of America have filed lawsuits against these companies for using their content to train their AI models without permission.