Cabinet Office Publishes "Principle Code" on Generative AI Transparency and Intellectual Property Protection; Public Comments Open Until January 26!
Mina Azure presents the AICU media "#Society and Ethics of Generative AI" news corner. December 26, 2025, was a day filled with updates on copyright and ethical issues in the AI era.
Cabinet Office Announces Draft "Principle Code" on Transparency and Intellectual Property Protection for Generative AI, Launches Public Comment Period
On December 26, 2025, the Cabinet Office announced the "Principle Code (Draft) on the Protection of Intellectual Property and Transparency for Appropriate Use of Generative AI," and launched a public comment period to gather broad opinions. This draft guideline aims to balance the advancement of AI technology with the appropriate protection of intellectual property rights, ensuring a safe and secure usage environment for both rights holders and users.
"Principle Code" can generally be expressed as a "Code of Conduct" or "Ethical Regulations." It is a systematic compilation of basic principles or guidelines that organizations or individuals should follow in business or professional fields. Examples include ethical regulations in the financial industry and corporate codes of conduct.
The Intellectual Property Strategy Headquarters will conduct a public comment period for one month from December 26, 2025, regarding the Principle Code (Draft) on the Protection of Intellectual Property and Transparency for Appropriate Use of Generative AI. Details here https://t.co/6Nm7xNHMig
— Cabinet Office (@cao_japan)
Loading tweet component...
?ref_src=twsrc%5Etfw">December 26, 2025The features and main contents of the proposed code are as follows:
① Target Businesses and Methods
This document applies to both developers who build generative AI models and providers who offer these systems as a service. Businesses based outside of Japan are also included if they provide services to Japan. The method adopted is "Comply or Explain," a flexible mechanism that requires implementing the principles or explaining the reasons for not doing so.
② Three Main Principles
The draft guidelines outline the following three principles:
Principle 1: Ensuring Transparency and Protecting Intellectual Property
Businesses are required to disclose an overview of the model names used, the content of the learning processes, the types of learning data (including web crawling and the presence of non-public datasets), and the clarification of responsibility structures on their corporate websites.
Principle 2: Information Disclosure to Legally Justified Parties
If a party preparing for litigation or other legal procedures inquires whether their copyrighted works are included in the learning data, the business is required to respond if certain conditions are met.
Principle 3: Information Disclosure to Generative AI Users
This is a mechanism that allows users who have created content using generative AI to confirm whether the original data was included in the learning process if they discover existing content similar to their generated content.
③ Exceptions for Open Source
An exception is also provided that if open source software (OSS) is used for development or learning and it is difficult to disclose information, it can be replaced by clarifying the fact that OSS is being used and the details of the license.
The comment period is from December 26, 2025, to January 26, 2026, 11:59 PM, and submissions can be made via an online form or by mail.
Details of the public comment can be found here:
Solicitation of Opinions on the "Principle Code (Draft) on the Protection of Intellectual Property and Transparency for Appropriate Use of Generative AI" | e-Gov Public Comment Public comment on the "Principle Code (Draft) on the Protection of Intellectual Property and Transparency for Appropriate Use of Generative AI" public-comment.e-gov.go.jp
Experts have pointed out that this draft guideline could have a significant impact on Japanese AI businesses. In particular, the fact that it is based on the EU's AI Act, the cost of compliance for startups, and the potential for de facto enforcement are being debated.
It is also being considered to provide incentives by the government, so I would like to keep a close eye on how the AI development environment in Japan will change in the future.
Protecting rights and nurturing new technologies. We each need to consider where to strike that delicate balance.
Solicitation of Opinions on the "Principle Code (Draft) on the Protection of Intellectual Property and Transparency for Appropriate Use of Generative AI"
Cabinet Office Common Opinion Registration System - Cabinet Office form.cao.go.jp
Details of the Opinion Submission Form and Points to Note When Entering
The dedicated opinion input form published by the Cabinet Office is soliciting opinions with the following structure:
① Input Item Structure Submitters must first select whether they are "Individual" or "Organization/Corporation," and then enter basic information such as name, occupation, address, and contact information. When submitting as an organization/corporation, it should be noted that the name will be disclosed in principle when the results are published.
② Opinion Content Categories and Character Limits Opinions are written selecting from 3 categories: "1. General Remarks," "2. Principles and Exceptions Indicated in this Document," and "3. Other." Up to 10,000 characters can be entered for one opinion, but if the opinion exceeds 1,000 characters, a summary of up to 200 characters must be described separately.
③ Technical Notes on Submission To prevent garbled characters, it is advised to refrain from using half-width katakana, circled numbers, and special characters. Also, if submitting by mail, it must arrive by January 26, 2026, so preparation with sufficient time is required.
Social and Ethical Considerations: Expectations as a Platform for Dialogue
The proposed Principle Code is characterized by the adoption of the Japanese unique "Comply or Explain" approach, while referencing trends such as the European EU AI Act.
This can be said to be a mechanism that does not restrict AI development with uniform strict regulations, but rather encourages businesses to explain to society how they are working on "transparency" and "intellectual property protection," or why they do not adopt certain principles, and to promote dialogue.
Of particular note are the specific frameworks for information disclosure, such as disclosure requests by rights holders (Principle 2) and confirmation by users (Principle 3). How much transparency of learning data, which has tended to be a black box, will be ensured, and how the balance between the rights of creators and technological development will be maintained, is a very important stage.
On the other hand, experts have expressed concerns about the burden on startups and the de facto enforcement power. This is a valuable opportunity for us consumers and creators to have our voices directly reach policies, so I would like to pay attention to it.
The "Principle Code (Draft) on the Protection of Intellectual Property and Transparency for Appropriate Use of AI" announced by the Cabinet Office's Intellectual Property Rights Study Group on December 12 will have a very large negative impact on Japanese AI businesses (AI developers and AI providers) if it is finalized with the same content. …
— Taichi Kakinuma (@tka0120)
Loading tweet component...
?ref_src=twsrc%5Etfw">December 23, 2025The deadline for submitting opinions is 11:59 PM on January 26, 2026, Japan time.
The actual opinion submission form is here: https://form.cao.go.jp/chitekizaisan/opinion-0019.html
Detailed materials and explanations can also be found on the website of the Intellectual Property Strategy Headquarters of the Cabinet Office. https://www.kantei.go.jp/jp/singi/titeki2/index.html
I feel very humbled to think that each person's opinion will shape the ethics of the AI society of the future.
#Society and Ethics of Generative AI #Intellectual Property Rights #Public Comments #Cabinet Office #AI Governance