Washington — The Computer & Communications Industry Association has filed comments on Japan’s draft “Principle-Code for Protection of Intellectual Property and Transparency for the Appropriate Use of Generative AI.” While CCIA supports the goal of promoting transparency as a core element of trustworthy AI, this draft’s prescriptive disclosure requirements risk undermining Japan’s innovation-friendly environment.
As noted in CCIA’s comments, granular transparency mandates, in addition to being burdensome, may expose proprietary trade secrets relating to AI training. Furthermore, requirements for source-level attribution remain technically infeasible for large-scale models that do not store discrete records. The proposal threatens to reopen settled copyright policies regarding lawful data analysis for machine learning, potentially discouraging investment and slowing AI adoption across Japan.
The following can be attributed to CCIA’s Vice President of Digital Trade, Jonathan McHale:
“Japan has long been a leader in fostering a permissive regulatory and legal environment for AI development, helping it attract significant investment in this burgeoning field. However, the draft Principle-Code risks imposing overly restrictive mandates that could create significant operational burdens and stifle the very innovation the government seeks to accelerate. We urge Japan to adopt a narrower scope for its disclosure framework, in alignment with flexible, risk-based international standards.”
