OpenAI: Impossible to create AI models without copyrighted material

The massive introduction of Artificial Intelligence and the related power of AI modelsresulted in several legal disputes regarding copyright.

According to what was stated by OpenAIa company he created and manages ChatGPTHowever, copyrighted material is essential for tools of this type. The statement was released in the context of an investigation by House of Lords of the United Kingdomin which the relationship between large language models (LLM) and material covered by copyright.

As is now well known the aforementioned ChatGPT, as well as GIVE HER and other similar tools, must necessarily be powered by external content in order to function. Most of the texts and images useful for this purpose are obviously the intellectual property of specific subjects (individuals or other companies). In fact, AI models scour the network collecting material that, at least theoretically, they could not exploit.

According to the document sent by OpenAI to the House of Lords “Because copyright today covers virtually every type of human expression – including blog posts, photographs, forum posts, snippets of software code, and government documents – it would be impossible to train today’s leading AI models without using copyrighted materials. copyright“.

OpenAI then wanted to explain how to limit the data for training AI models to books and images of public domainCreated more than a century ago” would not provide AI systems capable of meeting the needs of contemporary users.

The document in question follows the recent lawsuit brought by New York Times towards OpenAI and Microsoft. In this sense, the famous newspaper claims that there has been an abuse of its contents, with ChatGPT having drawn from its articles. OpenAI responded to the lawsuit on its website Monday, arguing that the lawsuit is without merit and affirming its support for journalism and partnerships with news organizations.

OpenAI’s defense is largely based on the legal principle of fair usewhich allows limited use of copyrighted content without the owner’s permission in specific circumstances.

This case is not the first in which the company directed by Sam Altman calls for fair use when it comes to data and AI training. In fact, last August a very similar situation occurred, with a lawsuit that also involved the comedian Sarah Silverman.


Please enter your comment!
Please enter your name here