According to the Nihon Keizai Shimbun, on June 4, the court sentenced Shota Mizuto to three and a half years in prison. Mizuto was charged with sharing illegal images of girls in a group chat with other teachers and possessing AI-altered sexual images of children.
The key issue was whether the AI-generated images constituted child exploitation material. Some of the images Mizuto stored were altered photographs of actual girls, transformed into sexualized images by AI. The court determined that these images could be perceived by the average person as being based on real children, thus qualifying as child exploitation material.
This ruling does not imply that all AI-generated sexual images of children will be prosecuted. The court emphasized that the composites were based on actual photographs of children. If an image is linked to a real child, it can be seen as a violation of the rights of the victim, even if it is not a direct photograph.
In Japan, there has been ongoing debate about whether sexual images created by generative AI fall under existing child pornography regulations. The current legal framework is predicated on the protection of real children, making the connection of AI images to specific children a central issue.
This ruling is expected to intensify discussions on regulating AI-generated sexual deepfakes in Japan. As generative AI becomes more prevalent, the increase in composite images that are not direct photographs raises questions about the legal responsibilities associated with AI-altered content based on real individuals.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.