Japan is on the brink of finalizing a controversial draft code regarding intellectual property (IP) and artificial intelligence (AI) that may undermine its reputation for fostering innovation. The code, developed by Japan’s Intellectual Property Strategy Headquarters, aims to establish a system for rightsholders to trace how AI models utilize their content during training. However, critics argue that the requirements outlined in the draft are technically unfeasible and counterproductive to Japan’s previously pro-innovation stance. This development comes as Japan and the United States, under the leadership of President Trump and Prime Minister Takaichi, seek to deepen economic and technological ties, emphasizing the significance of AI in their bilateral relationship.
Enacted in 2025, Japan’s AI Promotion Act fostered a regulatory environment designed to encourage innovation. Coupled with the Hiroshima AI Process, initiated during Japan’s G7 presidency, the act positioned Japan and the U.S. in alignment on many AI policy issues. However, the new draft code could jeopardize this alliance by imposing stringent disclosure requirements that many AI firms deem impractical. Under the proposed regulations, developers of generative AI systems are mandated to disclose the architecture and training methods of their models, detail sources of training data, and maintain comprehensive records of their systems over time. These obligations extend to any developer offering services in Japan, regardless of their headquarters, and non-compliance could lead to public explanations of their shortcomings.
The draft code faces significant criticism for its reliance on assumptions that do not hold for large-scale AI systems. For instance, confirming whether specific content appeared in training datasets is already challenging, and the requirement to determine if training data includes content “identical or similar” to generated outputs is deemed technically unfeasible for advanced models. These models typically learn through statistical relationships rather than retaining specific data points. The draft’s use of phrases like “to the extent technically possible and reasonable” suggests that firms may invoke these issues as bases for non-compliance, potentially leading to a registry filled with explanations rather than meaningful disclosures.
Moreover, the code introduces recordkeeping obligations without clear definitions of their scope or retention periods, leaving firms uncertain about compliance timelines. This ambiguity complicates the ability of companies to ascertain when they have fulfilled their obligations, thereby increasing the risk of public non-compliance findings. Additionally, the level of detail demanded in disclosures could compromise sensitive technical and commercial information, generating concerns about exposing proprietary methods to competitors and malicious actors.
During the public comment period, both domestic and international firms expressed concerns that these stringent requirements could deter legitimate data practices in machine learning. They warned that the draft code, instead of enhancing IP protections, could inadvertently weaken competitive advantages in the industry. The flaws inherent in the draft not only challenge its feasibility but also risk placing Japan at odds with a global consensus favoring balanced AI governance.
In contrast, the United States continues to advocate for a light-touch approach to AI regulation. The Trump administration has laid out a national AI policy framework that explicitly rejects heavy disclosure and traceability burdens. Even the U.S. Copyright Office, which previously expressed concerns about AI-related regulations, has not called for mandatory disclosure requirements akin to those proposed in Japan. This divergence could position Japan in opposition to its long-time ally just as both nations strive to assert leadership in the evolving AI landscape.
The Japanese government needs to reconsider the draft code before finalization. While the challenge of addressing IP in the context of generative AI is valid, the current requirements are unlikely to achieve their intended goals and may discourage firms from operating in Japan. A more viable solution would focus on establishing enforceable standards and promoting technically feasible transparency while aligning with international efforts already underway. Failure to revise the draft could jeopardize Japan’s position in the global debate over AI governance, where the balance between innovation and regulation remains critical.
See also
Tesseract Launches Site Manager and PRISM Vision Badge for Job Site Clarity
Affordable Android Smartwatches That Offer Great Value and Features
Russia”s AIDOL Robot Stumbles During Debut in Moscow
AI Technology Revolutionizes Meat Processing at Cargill Slaughterhouse
Seagate Unveils Exos 4U100: 3.2PB AI-Ready Storage with Advanced HAMR Tech
