Read in PDF version (including citations)
On December 26, 2024, the AI Strategy Council and the AI Institutional Research Group released the “Draft Interim Report”. From the perspective of striking a balance between promoting innovation and addressing risks, it suggested that, while continuing to rely on existing guidelines and other soft law measures, a specific legal framework should be implemented only for those cases where it would be difficult to expect businesses to take voluntary actions. In light of this, the Prime Minister Ishiba expressed his intention to move forward with the preparation of a new bill balancing AI innovation acceleration and addressing risks, which will be submitted to the Diet at an early date.
I. Past Trends.
International discussions on regulatory frameworks for AI are progressing, as demonstrated by the recent adoption of the EU AI Act, a comprehensive regulation enacted in the EU. In Japan, until now, rather than a comprehensive AI legislation, the competent authorities in each field have been addressing risks and issues associated with AI through laws and soft laws.
In July 2024, the AI Institutional Research Group was established under the AI Strategy Council with the aim of studying the ideal way to structure an AI system, including whether an appropriate regulatory system is necessary. Recently, the AI Strategy Council and the AI Institutional Research Group have published the “Draft Interim Report” (hereinafter referred to as the “Draft Interim Report”), which summarized the future direction of Japan’s AI policy and legal framework.
This article will outline the content of the Draft Interim Report and the status of the government’s actions based on it.
II. Summary of the Draft Interim Report.
1. Basic approaches to legal systems.
1.1 Balancing the promotion of innovation and risk mitigation.
The Draft Interim Report first set out the fundamental idea that an AI policy must balance the need to promote innovation and address risks. It then examined both the advantages and disadvantages of laws, regulations and soft law, pointing out that, given the high level of compliance awareness among Japanese companies, an excessive regulation may inhibit innovation. In order to ensure both the promotion of innovation and response to risks, the Draft Interim Report indicated the basic policy: laws and regulations and soft law, such as guidelines, should be appropriately combined and respect of the autonomy of businesses by limiting the use of regulatory measures to those regulations by law should be limited to cases where it is not possible to expect businesses to take voluntary action.
1.2 Promoting international cooperation.
There is an active discussion on AI governance in the context of an international framework, through initiatives such as theHiroshima AI Process and the OECD.In light of this, the Draft Interim Report suggested that the implementation of AI-related legal systems and measures should be based on the agreements or arrangements and concepts recognized and accepted in such international initiatives. It further expressed the need to ensure consistency and
interoperability between international and Japanese norms regarding safety and other aspects to be met, in order to facilitate the overseas expansion of Japanese businesses and access of Japanese citizens to AI services worldwide.
2. Direction of specific legal systems and policies.
In accordance with the above basic policy, the Draft Interim Report outlineed the direction of specific legal systems and policies , including: (a) strengthening the government’s function of a strategic leadership board; (b) taking legislative measures to ensure transparency and appropriateness in order to improve the safety of AI, and to enable the government to investigate and collect information on serious incidents; and (c) developing government procurement guidelines specific to AI.
2.1 Strengthening the government’s function of a strategic leadership board and formulating strategy.
While responses to individual risks are generally based on existing laws, there are also cases where an integrated and cross-cutting approach is needed for AI. Therefore, the Draft Interim Report pointed out the necessity to strengthen the government’s function of a strategic leadership board to overlook the entire process and to formulate a strategy or a basic plan. In this regard, it was suggested to strengthen the function of a strategic leadership board for AI to clarify its authority when requesting cooperation from related administrative agencies.
2.2 Improving safety.
The Draft Interim Report presented the following points regarding the initiatives needed to ensure transparency and appropriateness for improving the safety of AI.
i. Ensuring appropriateness and transparency through AI life cycle:
・ To ensure appropriateness, it is advisable for the government to establish guidance based on the purpose of international rules such as the Hiroshima AI Process, and to encourage businesses to voluntarily comply with them.
・ With regard to ensuring appropriateness, including ensuring transparency, the government should ascertain the situation of business operators through surveys, and provide the necessary support in accordance with the results obtained from these surveys, including measures based on existing laws and regulations. In this context, as the government’s understanding of the situation of business operators and the provision of the necessary support cannot be achieved without the cooperation of business operators, it is appropriate to take action through the legal system so that information sharing and cooperation from domestic and foreign business operators can be sought.
ii. Strategic promotion related to safety evaluation and certification practiced by domestic and overseas organizations:
・ The Draft Interim Report also suggested that the implementation of safety evaluation and third-party certification system would be effective means.
iii. Investigation and information dissemination by the government regarding serious incidents:
・ As AI has been developing rapidly in recent years, and various risks are increasing, it would be appropriate for the government to first gather information and understand the actual status regarding the development, provision, use of AI, and then provide information to the citizens to the necessary extent, while taking into account confidential information.
・ In the event of a serious accident resulting from the use of AI, the government should take steps to prevent its occurrence or escalation, and also to call the attention of business operators developing and providing AI to measures preventing the recurrence of such accidents.
・ Furthermore, since the survey and dissemination of information cannot be accomplished without the cooperation of business operators, it would be appropriate to take action through legal systems so that the government can require domestic and overseas business operators to cooperate to provide information.
2.3 Developing government procurement guidelines specific to AI.
It was also pointed out that the rate of AI use by individuals and companies in Japan is significantly lower than in other countries and suggested that the government should take the lead in using AI to promote its use by citizens in order to ensure international competitiveness.
The Draft Interim Report noted the importance of developing government procurement guidelines specific to AI, as well as deepening the existing AI-related guidelines.
III. The policy direction for legislation based on the Draft Interim Report.
The Draft Interim Report is open for public comment until January 23, 2025. In response to it, the Prime Minister Ishiba has outlined the following policy:
・ The government will move forward with measures to ensure that a new bill that balances the need to accelerate AI innovation with the need to address risks can be submitted to the Diet at an early date.
・ An “AI Strategy Headquarters” consisting of all cabinet ministers will be established to strengthen the government’s function of a strategic leadership board for the AI policy.
・ Guidelines for procurement and use of AI will be developed. Ministries and agencies, as well as local authorities, will assess the actual state of AI implementation in infrastructure and other areas, and take necessary measures such as revising the guidelines.
・ The government will develop guidelines in accordance with the “Hiroshima AI Process” to encourage compliance by the private sector, while also taking necessary measures such as investigating malicious cases and collecting information from AI developers.
・ The government will also support the development of technologies to tackle fake information and the formulation of standards and evaluation methods for the safety of AI.
IV. Summary.
As explained above, based on the Draft Interim Report of the AI Institutional Research Group, AI governance in Japan should be left to voluntary efforts of businesses through soft law such as guidelines, while the government’s function of a strategic leadership board and its authority to investigate and collect information should be strengthened.
Thus, a new legal system is expected to be introduced in Japan to address the risks of AI, and businesses should closely monitor future policy trends.
Moreover, according to the Draft Interim Report, the new legal system is expected to be positioned as a complement to efforts based on existing frameworks such as guidelines. Therefore, it is necessary for businesses to continue to address risks surrounding the use of AI through measures based on Japan’s existing systems and guidelines, such as the “AI Guidelines for Business Ver1.01”. Also, depending on the scope of the business, it may be necessary to take action in line with international trends, such as the EU AI Act.
［AIL AI Newsletter Series.]
* For more information on trends in conventional policies and other initiatives in Japan, please click here.
* For more information on the “EU AI Act”, please click here.
