Japan’s Newspaper Association Concerned over Generative AI

The Yomiuri Shimbun

20:00 JST, May 17, 2023

The Japan Newspaper Publishers and Editors Association has released a statement expressing its concern about the unauthorized and unregulated use of news content by generative artificial intelligence. The association has demanded that the government hasten institutional measures to make generative AI harmonize with society from the perspective of the entire legal system.

In its opinion on the use of news content by generative AI, released Wednesday, the association said that unauthorized and unregulated use by generative AI of articles, photos, images and other content of news organizations could harm the healthy discourse space. Hence, the association stressed, “If such disruption of the discourse space proceeds, it will lead to social disturbance.”

To address the situation, the association also urged the government to establish relevant laws and help social systems to keep up with the developments, in addition to creating rules requiring AI developers to disclose which news organizations’ content has been used for training data.

First and foremost, the association explained that AI can generate answers that often contain mistakes and inaccurate information, and therefore, it can spread disinformation, harmful information or politically motivated information to mislead public opinion on the internet. The association also expressed concern that the technology also makes it possible to mass-produce new articles by combining multiple articles from news companies. The association also noted that data used for AI training is likely to include “sensitive personal information,” which, in principle, is prohibited from even being obtained without the consent of the individual concerned, and there is a risk that AI may include such information in its answers.

The association then noted that there is growing concern about unauthorized and unregulated use by AI of articles, photographs, images and other content of news organizations such as newspapers and news agencies. Specifically, the association said AI is importing large quantities of articles, photos, images and other data from news organizations’ digital news sites without permission and developing commercial AI services for third parties. The association claimed such a practice “could conflict with the existing database sales market and unreasonably prejudice the interests of the copyright owner.”

The association also noted that an amendment to Japan’s Copyright Law in 2018 legalized in principle the collection and use of existing copyrighted works without permission in the process of developing AI and other technologies. Such a rule is significantly relaxed compared to the standards in Western nations.

Regarding the deliberation process for the amendment bill, the association said: “As there were only a few references to AI, there was no indication from the government that AI could generate new expressive work and threaten right holders. So the right holders did not believe that the use of copyrighted works for technological development would become a problem, and for this reason, the law was amended to favor AI development without major debate in the Diet.”

The association stressed: “If such disruption of the discourse space proceeds, it will lead to social disturbance. It cannot be overlooked in terms of protecting democracy.”

The association called on the government to take urgent institutional measures to ensure the harmonious development of generative AI with society from the perspective of the entire legal system, including the Copyright Law and the Protection of Personal Information Law.