AI Regulations in Japan Under Development; Calls for Reform of Copyright Law by Publishers

Reuters file photo
Illustration taken on June 23.

Effective regulations for domestic artificial intelligence providers are still under development and Prime Minister Fumio Kishida said, “It is important to establish international governance that promotes the benefits of AI while appropriately addressing the risks.”

Economic security minister Sanae Takaichi told reporters at the press conference Friday, that AI development “must be balanced with the need to ensure safety and to promote the spread of AI.”

The Group of Seven nations at the end of October agreed on international guidelines and an international code of conduct for AI developers as part of the Hiroshima AI Process led by the chair, Japan. Those guidelines call for appropriate measures to be taken from the pre-market stage, illustrating risks such as the development of weapons using AI and the promotion of disinformation, prejudice and discrimination.

In line with those international guidelines, the government drafted its guidelines for domestic AI providers in November, which includes a discussion about the implementation of a third-party certification system. The government plans to proceed with the design of a detailed regulatory system. As it is unclear whether the system will serve as a deterrent to AI operators who try to develop products without due regard for safety, the government is reluctant to establish penalties for violations.

Regulations governing operators are also controversial in other countries. The European Union aims to pass an AI law by next spring that will strongly regulate operators, including with the use of penalties. However, some member countries, particularly France, Germany, and Italy, are of the opinion that self-regulation by AI developers is sufficient. These countries are concerned that the new law would hamper the activities of their domestic AI providers, and talks over AI regulation within the EU may be protracted.

In Japan, the focus is also on the state of Copyright Law, which currently permits AI to learn without the permission of copyright holders. A subcommittee of the Council for Cultural Affairs in the Cultural Affairs Agency is discussing the issue with the aim of finalizing measures by the end of March. The Japan Newspaper Publishers & Editors Association and others are calling for stronger measures, including legal reform.

The originator profile, a digital technology that clearly identifies the sender of information, is considered to be an effective countermeasure against disinformation.