Anthropic Innocent In Cases Of Using Books To Train AI

Anthropic won the class action lawsuit filed by the author (photo: dock. Anthropic)

JAKARTA Anthropic, an Artificial Intelligence (AI) company, has settled a class action lawsuit with the author. This lawsuit relates to the use of authors' works to train AI models.

The case, known as Bartz vs. Anthropic, has been running since the end of last year. This lawsuit was filed by the author on the basis of unlawful use of fiction and non-fiction books or without applying for an official permit.

According to the authors, actions like this violate their copyright. After a year in progress, the appeals court decided to calm Anthropic and stated that authors' books were 'natural use'.

This means that Anthropic did not make a serious mistake by using commercial books to train his AI model. Even though he is considered not to have made a mistake, Anthrop still receives financial sanctions due to material from his book.

Instead of using original books, many books contain pirated material. It was not explained how many books. Launching from TechCrunch, the court also did not share details of the settlement agreement for the case.

Anthropic and the author's legal representative also did not comment on the court's decision. Therefore, it is not yet known how much the financial value of the settlement is.

Previously, Anthropic had praised the court's decision regarding the 'natural use' of authors' books. According to him, this decision is a big win for the generative AI model. This shows that the data in the book can be used to train AI.

"We believe it is clear that we obtained a book for only one purpose building a large language model and the court clearly states that its use is reasonable," said Anthropic. This statement shows their confidence in their legality.