In January 2023, the Liberal Democratic Party of Japan launched a project team (“AIPT”) on the evolution and implementation of AI, led by Masaki Taira, a member of the House of Representatives, to consider Japan’s AI strategy and policy recommendations. Based on these considerations, AIPT member’s a private proposal for new legal governance in our country, specifically targeting particularly powerful generative AIs (called “frontier AI models”). has been submitted to AIPT and published as the “Basic Law for Promoting Responsible AI”1. The Draft AI bill will regulate developers of foundation AI models of a certain scale and purpose.
The Draft AI bill aims to balance risks and benefits brought about by AI through the introduction of appropriate AI governance. Specifically, it seeks to restrict risks of infringement on the rights and interests of citizens, while maximizing the benefits of healthy AI development, including AI‐driven innovation. This is to be achieved by public‐private joint regulation against developers of specific foundation AI on sharing of risk information, establishment of a governance‐system, disclosure, and so on. The bill specifically targets so‐called “specific AI foundational models” with
significant social impact, known as frontier AI models. It does not intend to include…
