JAPAN Law and Practice Contributed by: Hiroshi Mitoma, Tomohiko Iwasaki, Kosuke Hamaguchi and Akira Komatsu, Nagashima Ohno & Tsunematsu
8. Artificial Intelligence 8.1 Board Oversight of AI
set up. On the other hand, at companies where AI utilisation has not progressed as far, there are many cases where no dedicated department overseeing AI utilisation exists, and departments such as corporate planning, DX, legal and compliance are involved in AI utilisation only to the extent relevant to their respec - tive functions. 8.3 Liability Exposures Arising From AI Use Principal Risks The principal risks include copyright risk, reputational risk and tort liability. With respect to copyright, as a general rule, if an AI-generated output is found to have similarity and dependence with existing copyrighted works, it will be considered a copyright infringe - ment. A generative AI that produces output similar to existing copyrighted works poses a risk of copyright infringement. For example, characters appearing in Japanese animation were generated in connection with OpenAI’s Sora2, which resulted in the Cabinet Office requesting that OpenAI refrain from committing copyright infringement. With respect to reputational risk, given that there is criticism in certain industries concerning the use of generative AI, there are risks such as the discontinuation of services or withdrawal of advertisements utilising generative AI. Furthermore, with respect to damages arising from AI generating erroneous content, there is the question of which AI business operator (developer, provider or user) bears what type of liability. In principle, the mat - ter is governed by the principles of tort law. In April 2026, the Guidance on the Interpretation and Applica - tion of Civil Liability in AI Utilization [Version 1.0] was published, which attempts to organise what liability arises in various cases. In addition to the liabilities above, using AI may cause exposure to the risk of breach of confidentiality if inputs include confidential information provided under a non-disclosure agreement, and the risk of violation of the Act on the Protection of Personal Information (APPI) if using personal data in inputs does not comply
In Japan, no special requirements are imposed on companies regarding board composition, committee mandates, or risk and control matters, in connection with the use of AI. In Japan, the Act on Promotion of Research and Development, and Utilization of Artifi - cial Intelligence-related Technology (the “AI Promotion Act”) was enacted in 2025. The principal provisions of the AI Promotion Act applicable to companies are (i) the obligation to co-operate with the measures implemented by the national government and (ii) the government’s authority to provide guidance, advice, information and take other necessary measures to companies based on the results of government inves - tigations into infringement of citizens’ rights and inter - ests. The AI Promotion Act does not regulate matters relating to board oversight in connection with com - panies’ use of AI. 8.2 AI Use-Related Risks Governance Frameworks and AI Governance The AI Guidelines for Business (the “Guidelines”) provide AI business operators (developer, provider and user) with fundamental principles regarding the risks and countermeasures associated with utilising AI. Among other things, the Guidelines highlight the importance of AI governance. The Guidelines intro - duce that, in order to respond to complex and rapid changes, it is important to practise “agile govern - ance” in which a cycle of “environmental and risk analysis”, “goal setting”, “system design”, “opera - tions” and “evaluation” is continuously and rapidly iterated among multiple stakeholders. Version 1.0 of the Guidelines was formulated in April 2024 but the Guidelines continue to be updated: Version 1.2 was published in March 2026. Corporate Bodies or Functions Responsible for AI Strategy In Japan, there are various corporate bodies responsi - ble for AI strategy and risk management. At companies where the promotion of AI utilisation has advanced, there are cases where a position such as CAIO (Chief AI Officer) has been established within the manage - ment to handle AI strategy and risk management, or where a department overseeing AI utilisation has been
with the APPI. Enforcement
Pursuant to the AI Promotion Act, the AI Strategy Headquarters has been established within the Cabi -
395 CHAMBERS.COM
Powered by FlippingBook