Media Organizations in West Sue Generative AI Firms Over Copyright Infringement; Organizations in Japan Likely to Follow Suit

Reuters file photo
A Perplexity logo is seen in this illustration taken in May 2024.

There have been a series of cases in which media organizations in the West have sued generative artificial intelligence companies for copyright infringement over using their articles without permission. There are already more than 30 such news organizations, and many of them are located in the United States.

In Japan, it has been pointed out that there are issues with the Copyright Law, as it fails to prevent generative AI companies from using news articles for free to train their AI systems. It is likely that more media organizations in Japan will take their cases to court.

Generative AI services, such as OpenAI’s ChatGPT, can be trained, as well as gather information from various online sources, allowing it to answer users’ questions. Therefore, stories written by media organizations are vital for generative AI firms to improve the accuracy of their products, as the articles are thoroughly fact checked.

If AI companies obtain permission from the media organizations to use the stories as a study tool or reference, it would not be an issue.

However, many are concerned that the stories have been used without consent.

According to British media industry journal Press Gazette, there were at least eight cases as of July in which AI development companies were sued for using stories without permission. The defendants include OpenAI, Microsoft Corp. and Perplexity, all U.S. companies.

In one of the cases, The New York Times filed a lawsuit against OpenAI and Microsoft, which are affiliated with each other, in December 2023. The lawsuit claimed that several millions of articles had been used for AI training and other purposes without permission, resulting in several billions of dollars in damages.

Specifically, the newspaper claimed that when ChatGPT was asked to provide what was written in a New York Times article, which was not available for free, about an avalanche in the state of Washington in 2012, the generative AI included parts from that very article almost word for word.

There are also cases in which generative AI companies become affiliated with media organizations and make payments to them. It has been reported that Amazon.com, Inc. will pay up to $2.5 million (about ¥3.7 billion) annually to The New York Times to use their articles for generative AI services. However, Press Gazette reports that there have been only 30 or so such cases.