The Agency for Cultural Affairs announced the results of public comments on the “Draft Approach to AI and Copyright. The number of opinions submitted amounted to 24,938, of which 73 were from corporations and organizations. In addition, the February 29, 2024 version of the “Draft Concept” based on the public comments was also released.
Subcommittee on Legal System of the Copyright Subcommittee of the Council for Cultural Affairs (7th) | Agency for Cultural Affairsaipubcom
Example
Generate AI regulatory laws. Humans and AI should be considered separately, as in the case of automobile traffic laws. Generating AI surpasses humans in quantity and reproducibility, but even if it continues to evolve, it will remain asleep like a plant because it has no life (desire for predation) and will not develop an ego, so it is a tool similar to a copy machine, and tools do not need copyrights. As an example of a law regulating generated AI, it would be OK to learn AI, but it would be NG to publish it on the Internet, etc., plus a license system, etc. This would not hinder AI development and would be compatible with copyrights. I know there is a fine and difficult line to draw, but I hope you will stick to it and take action! Or should we clarify “cases in which copyright can be specifically disregarded” in Article 30-4 Currently, there are many people who interpret “it is OK to ignore the basic copyright school regarding generated AI”, and it is very difficult to understand. If the current situation is left to the judiciary, the victims will cry themselves to sleep, or the court will not have enough to deal with. It is no exaggeration to say that regulations are being implemented in other countries and only Japan is in a state of inaction. If the situation continues as it is, various content industries will no longer be traded abroad as they are tainted by AI, people will not be trained because AI will take away their jobs, industries will die, the economy will stagnate, the birthrate will accelerate, and Japan will perish. Please help Japan!
Below is one of the public comments. Can you extract anything useful from this text?
The following useful points can be extracted from this public comment
- Need for a law regulating generative AI: Argues that the differences between generative AI and humans should be clarified and regulated by law.
- Characteristics of generative AI: Opinion that generative AI surpasses humans in quantity and reproducibility, but does not need copyright because it has no ego.
- Specific regulatory proposals: Proposal to allow up to learning, and to require a license to publish on the Internet, etc.
- Compatibility with copyright: Need for regulations that do not inhibit AI development and are compatible with copyright.
- Issues with current law: concern that current law is ambiguous and is interpreted as “basically ignoring copyright with respect to generated AI”.
- Trends overseas: A sense of crisis that regulations are progressing overseas and Japan’s response is lagging behind.
- Future Impact: Warns that without proper regulation, there will be serious consequences, including a loss of confidence in the content industry and economic downturn.
The above points will be helpful in discussions and policy making regarding the regulation of generative AI.
This page is auto-translated from /nishio/AIと著作権に関するパブリックコメント using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.