A proposal to regulate generative AI at the international level has been prepared for the first time. The Japanese government outlined a draft for the “Hiroshima AI Process,” a cabinet-level forum for discussing generative AI technology. It’s expected to serve as a basis for the world’s richest Group of Seven(G7) countries to discuss in detail this upcoming September.
The draft is aimed at preventing copyright infringement and the spreading of false information by inserting a digital certification into the edited version including news content, company websites, and advertisements.
The edited content contains a mark that features a digital signature, the so-called “Originator Profile,” which includes the name, company details, address, etc. of the original creator, with a click of a button. The digital signature is envisioned to follow the content when it’s shared on social media and other online sites.
This policy is to be supported by the development of a new web standard technology that can detect whether the content is AI-generated, according to the Yomiuri Shimbun. Combined with the Originator Profile, the new web standard technology is expected to play a vital role in preventing the use of false data for training generative AI models.
The international framework is being devised as countries around the world are trying to implement their own set of AI regulations. Japan, which is expected to release a set of laws regarding generative AI by the end of this year, is preparing to take a soft approach to not discourage development.
Meanwhile, the EU passed a landmark law in June that applies the strictest rules by dividing the use of AI into various risk categories. For uses of the technology labeled “high risk,” companies would have to go through multiple rounds of review for approval before their commercial release.