The Japanese Cabinet Office released a draft guideline on August 18 aimed at regulating generative AI development. The draft requires developers to disclose training data sources and collection methods, ensure transparency about copyrighted materials used in training sets such as comics music and movies, and set up internal notification systems to alert users when generated content mirrors protected works. It also urges developers to avoid using pirated source material. Though not legally binding the guideline aims to balance IP protection for creators with technological innovation support while narrowing the scope for AI companies to use opaque training processes in Japan's copyright sensitive content sector。