Partner Perspectives: Hiroyuki Tanaka

About our speaker:
Hiroyuki Tanaka is a partner of Mori Hamada & Matsumoto, admitted to practice in Japan (Daini-Tokyo Bar Association) and New York. His practice areas are data protection, IT and IP. He has extensive experience advising foreign clients on Japanese data protection law. He has extensive experience with various IP and technology disputes.

His practice area includes legal issues relating to AI (especially generative AI) and the protection of cybernetic avatars. He is an adjunct project professor at Keio University Graduate School of Law. He is also a member of the Cyber-Physical Sustainability Center of Keio University Global Research Institute (2023-present).

Could you please provide an introduction about yourself, and a brief overview of your law firm?

I am a partner at Mori Hamada & Matsumoto. I’m also a project professor at Keio University Graduate School of Law. Mori Hamada & Matsumoto, is one of the big four law firms in Japan and my practice areas are data protection, IP and IT. Recently, I have done a lot of work regarding legal issues relating to AI especially Generative AI.

AI is the global hot topic right now. Has Japan developed a national strategy for artificial intelligence? 

The Japanese government is very favourable to the expansion of AI. In April 2024, the ruling party, the Liberal Democratic Party, published a paper titled “AI White Paper 2024 Stage II New Strategy—Towards the World’s Most AI-Friendly Country.”

The paper states that Japan aims to become the country with the greatest understanding of AI and the most accessible environment for AI research, development and implementation globally.

Has Japan implemented rules or guidelines (including voluntary standards and ethical principles) on artificial intelligence?  

Yes, it’s a Ministry of Internal Affairs and Communication and the Ministry of Economy, Trade and Industry of the Japanese government jointly released AI Business guidelines in April 2020.

This guideline is not legally binding hard law but rather soft law.. The guideline covered items that apply commonly to AI developer, providers, and users as well as items that apply to each business operator individually, touching on various tasks.

Attachments include checklist and conceptual virtual case studies. This guideline is expected to be referenced in practice.

Given its prevalence, what are the main issues to consider when using artificial intelligence systems in the workplace? 

Many Japanese companies are using generative AI. There are some risks to the use of generative AI. When inputting information, there are risks, such as inputting information that raises concerns about, data protection laws, privacy and confidential information, and inputting information that’s raises concerns about infringement of copyright and other intellectual property rights.

And when outputting information, there are risks, such as incorrect information due to hallucinations , and outdated information based on the timing of machine learning.

Additionally, there are civil and criminal liability risks for outputting information that infringing and copyright and other intellectual property rights, violates privacy rights, constitutes illegal acts, such as defamation , or otherwise is unlawful or improper.

Furthermore, due to automated generation by generative AI, there is a risk that the company may not be able to obtain copyrights and other intellectual property rights. It is also important not to violate the terms of service of generative AI services.

Finally, not only legal issues but also ethical issues are important. Problems such as bias and the possibility of human manipulation can be problematic.

Please describe any civil and criminal liability rules that may apply in case of damages caused by artificial intelligence systems.

In March 2024, the “Perspectives on AI and Copyright” report was published by the Copyright Subcommittee of the Council for Cultural Affairs.
Approximately 25,000 opinions were submitted during the public comment process through the draft report, which shows people’s high level of interest in this issue.

The report organises the perspectives on globally recognised issue such as copyright and machine learning for AI, copyright infringement by AI generated content and attribution of copyright for automatically generated content by AI.

This report is neither low nor judicial precedent, but it presents specific perspectives on important practical and academic points regarding AI and copyright.

In Japan, there are no abstract provisions for so-called fair use, but there are individual copyright limitation provisions that anticipate machine learning by AI. A famous copyright scholar in Japan described Japan as a machine learning paradise. However, with the rapid expansion of generative AI, there are strong concerns, particularly among creators and media, about the inability to exercise copyright over extensive machine learning, and there are disputes over the interpretation of this clause depending on one’s position.

What privacy issues arise from the use of artificial intelligence? 

First and foremost, there is the issue of correcting personal information when developing AI. In particular, the question of how to clear the rule under Japan’s Act on the Protection of Personal Information, which requires consent in principle for collecting sensitive personal information, is becoming a problem.

The Personal Information Protection Commission issued an alert to Open AI in June 2023 regarding this issue.

Secondly, there is an issue of inputting personal information into AI.

Specifically, the extent to which the purpose should be carefully explained when inputting personal information into AI and profiling individuals is practically important issue.

Furthermore about sending personal data to AI service provider, the question of how to clear the requirement under Japan’s Act on the Protection of Personal Information, which stipulates that consent of data subjects is necessary in principle when providing personal data to a third party, as well as the cross-border transfer regulations, is becoming an issue. Evaluating the data processing agreement provided by the AI service providers under the Japanese law is required when conducting this examination.

How do you typically help to advise clients on matters around AI? 

For example, I provide advice on guidelines for companies’ AI usage. I also frequently receive consultation from companies regarding offering services such as application incorporating generative AI. In these cases, I offer advice on such issues as creating terms of use and privacy policies for users.

I also have opportunities to handle inquiries from global companies regarding cross sector legal regulations of various countries, being responsible for examining the Japanese portion. Additionally, I frequently receive inquiries from Japanese companies about regularly trends in foreign countries. Therefore, I try to keep abreast of international trends in AI regulations.