To use this site, please disable the ad blocking feature and reload the page.
This website uses cookies to collect information about your visit for purposes such as showing you personalized ads and content, and analyzing our website traffic. By clicking “Accept all,” you will allow the use of these cookies.
Users accessing this site from EEA countries and UK are unable to view this site without your consent. We apologize for any inconvenience caused.
Yomiuri Shimbun file photo The Cultural Affairs Agency in Kamigyo Ward, Kyoto.
The Yomiuri Shimbun
7:00 JST, November 22, 2023
The Cultural Affairs Agency has come up with a draft outline specifying issues that need to be examined in connection with copyright and generative artificial intelligence.
The agency submitted the draft outline to a subcommittee of the Cultural Council on Monday.
It has been pointed out that media outlets’ news content and the works of creative people are being used to train AI systems that generate images and texts — without the consent of the original materials’ copyright holders.
By the end of fiscal 2023, the agency plans to form a consensus about what actions would constitute cases that unreasonably prejudice the interests of copyright holders under the Copyright Law.
When the Copyright Law was revised in 2018, Article 30-4 was established to allow copyrighted works to be used for training AI without permission from the rights holders.
Even though the provision states that, in exceptional cases, copyrighted materials should not be used in this way “if the action would unreasonably prejudice the interests of the copyright owner,” no specific examples have been provided to illustrate what might constitute unreasonable infringement of the copyright holders’ interests.
As one point that future discussions need to focus on, the draft outline touches on the question of what actions should be considered exceptional cases.
The draft outline also presented the following questions as points that need to be addressed:
■ If an AI-generated work resembles an existing copyrighted work, how should it be determined whether the AI work infringes copyright?
■ In cases of copyright violation, how far could the use of the work in question be limited through demands for injunctions and other means?
■ To what degree must specific instructions be given to AI in order for an AI-generated work to be recognized as a copyrighted work?
The Japan Newspaper Publishers and Editors Association is calling on the government to expedite efforts to consider revising the Copyright Law and establish relevant rules.
The association argues that if unauthorized and unregulated use of copyrighted works continues for training AI, it will deal a serious blow to media outlets and thus affect people’s right to be informed.
“We think the draft outline also addressed the concerns expressed by the newspaper association,” said a Cultural Affairs Agency official. “We will examine individual points at issue.”
