Japan’s Cabinet Office proposed a revised Principle-Code on August 18 to set intellectual property (IP) protection standards for generative artificial intelligence (AI) businesses. The draft code emphasizes avoiding crawling pirate sites, respecting paywalls, and increasing transparency about IP risk management. It applies to both Japanese and foreign generative AI providers whose services are available in Japan, according to medianama.com.
The revised code follows a public consultation held from December 26, 2025, to January 26, 2026, and uses a ‘comply or explain’ approach. Under this framework, generative AI businesses must either adhere to the principles or publicly explain why they do not. The proposal aims to provide rights holders with more information about AI models, training data, and data-collection methods while maintaining Japan’s existing copyright laws.
This move addresses growing concerns over IP protection in the AI sector, especially regarding the use of copyrighted content for training AI models. By increasing transparency and setting clear guidelines, Japan seeks to balance innovation with rights-holder protections. The approach contrasts with more prescriptive regulations elsewhere, relying on voluntary compliance with public accountability.
The revised Principle-Code was discussed by the Study Group on Intellectual Property Rights in the AI Era on August 18. It represents Japan’s latest effort to create a governance framework for generative AI that covers both domestic and international companies operating in the Japanese market.