Japan's Current Status on Generative AI and Copyright: A Summary of Developments, Current Situation, and Key Issues

The Japanese House of Councillors published a report outlining the history, current legal perspectives, and key issues surrounding generative AI and copyright. The report includes a review of developments in Japan, other countries, and the private sector.

Copyright Law Revision

• In April 2017, the Copyright Subcommittee of the Council for Cultural Affairs (an advisory body for the Minister of Education, Culture, Sports, Science and Technology and the Agency for Cultural Affairs) proposed revisions to the Copyright Act. The proposed changes aimed to introduce flexible copyright limitations to address technological advancements, such as AI development.
• The Subcommittee proposed that, for non-commercial purposes such as machine learning for AI development, the use of copyrighted works should be considered distinct from the original purpose of the work and not generally harm the rights holder’s interests. This proposed a more flexible legal framework for such activities.
• In May 2018, the revised Copyright Act was passed.

Key Developments Post-Revision

• May 2023: The AI Strategy Council published preliminary conclusions on AI-related issues.
• June 2023: The Intellectual Property Strategy Headquarters announced the Intellectual Property Promotion Plan 2023.
• March 2024: The Legal System Subcommittee of the Copyright Subcommittee of the Council for Cultural Affairs published its perspective on AI and copyright.
• April 2024: The Agency for Cultural Affairs established a network of stakeholders to address AI and copyright issues.
• May 2024: The Intellectual Property Strategy Headquarters AI Era Intellectual Property Rights Review Committee published an interim report.
• June 2024: The Intellectual Property Strategy Headquarters announced the Intellectual Property Promotion Plan 2024.

Key Issues

• Development and Training Phase (Copyright Law Revision):
  • Will Article 30-4 of the Copyright Act promote AI development and innovation?
  • Is Article 30-4 of the Copyright Act appropriate in light of the purpose of copyright law?
• Generation and Use Phase (Disclosure Obligations for Training Data):
  • While the current Copyright Act does not mandate disclosure of training data for generative AI, it does allow refusal to disclose if there are valid reasons.
  • Some argue that this could put Japanese AI businesses at a disadvantage internationally.

Historical Information on Japan's Generative AI and Copyright:

• Early AI Developments: Japan's interest in AI predates the recent boom, with research and development efforts in robotics and machine learning emerging in the 1970s and 1980s.
• Early AI and Copyright Discussions: Discussions on the intersection of AI and copyright law in Japan began in the early 2000s, focusing on the use of copyrighted material for AI training and the implications for creative industries.
• Influence of International Trends: Japan's approach to generative AI and copyright has been shaped by international developments, particularly the discussions on copyright limitations for AI training in the European Union and the United States.
• Focus on Economic Competitiveness: The Japanese government's emphasis on promoting AI development in the 2010s, as part of its "Society 5.0" vision, highlighted the need for a clear legal framework that balances innovation with copyright protection.
• Public Awareness and Education: The government has taken steps to raise public awareness and educate stakeholders about the ethical and legal implications of generative AI through workshops, conferences, and online resources.
• Public Opinion and Ethical Considerations: Japanese society, while embracing technological advancements, has also expressed concerns about the potential impact of AI on creative industries, artists' livelihoods, and the integrity of cultural heritage.

Key Takeaways: Generative AI and Copyright in Japan

• Japan's copyright law has been revised to accommodate AI development. This includes a flexible framework for non-commercial use of copyrighted materials for AI training.
• There is ongoing discussion about balancing AI development with copyright protection. This includes debates about the appropriateness of current legal frameworks and the potential need for further revisions.
• Transparency around training data used for generative AI is a key issue. While the current law doesn't require disclosure, some argue that this could hinder Japanese AI businesses in a global market.
• Multiple government bodies and stakeholders are actively engaged in addressing AI and copyright issues. This includes the formation of new committees and the publication of reports and strategies.