Recently, a group of publishers and authors jointly filed a class-action lawsuit against Google, accusing the tech giant of using their copyrighted works without authorization in training its AI platform Gemini. The lawsuit includes well-known publishers such as Hachette, Cengage, and Elsevier, as well as authors Scott Turow and the S.C.R.I.B.E. organization.

The lawsuit documents state that Google not only used these copyrighted works in training AI but also deliberately removed or altered related copyright information to hide the fact that it had "stolen" the material. Although there are numerous copyright lawsuits against AI companies, two preliminary rulings in California have supported these tech companies, stating that using copyrighted works for AI training is considered "fair use" under U.S. copyright law. However, the company Anthropic was fined $1.5 billion for infringement, setting a record for the largest copyright penalty in U.S. history.

In this case against Google, the relationship between publishers and Google is relatively complex. The lawsuit states that the long-term collaboration between publishers and Google began with the Google Books project, which aimed to make books searchable rather than allowing users to view entire books. Google illegally copied works from these limited programs and used them for AI training without permission.

According to the complaint, internal Google documents mentioned that using copyrighted books for AI training could cause "serious issues" and potentially lead to fines of "hundreds of billions of dollars." Although Google has not yet responded to this matter, this case will undoubtedly have a profound impact on future AI training and copyright legal disputes.

Currently, although many copyright lawsuits involving AI are still being heard, the ruling in this case will provide new perspectives and interpretations for the legal community.