Post by : Bianca
Haleem
AI-generated videos that echo Studio Ghibli's signature style have attracted global attention and prompted pushback in Japan. CODA, a trade body representing major publishers including Studio Ghibli, has formally asked OpenAI to stop using copyrighted Japanese works in its AI training.
Sora 2, OpenAI's video-generation tool introduced on 30 September, quickly drew users for its ability to create Ghibli-like imagery. That similarity has sparked legal and ethical concerns. In a letter dated 27 October, CODA said many outputs "closely resemble Japanese content or images," indicating the AI may have been trained on protected material.
Under Japan's copyright law, using creative works without prior permission can amount to infringement. CODA criticised OpenAI's opt-out approach — which removes material only after it has been used for training — saying it conflicts with Japan's legal requirement that creators consent beforehand.
CODA set out two main demands: that OpenAI cease using members' works without explicit permission and that the company engage seriously with members' copyright complaints. The organisation emphasised the need to balance technological progress with the rights of creators.
The dispute follows other contentious Sora 2 incidents. OpenAI temporarily halted AI-generated images of Martin Luther King Jr after the estate raised objections over disrespectful portrayals, underlining growing scrutiny of the platform's content controls and ethical standards.
The episode adds to an expanding international debate over AI and copyright, with governments and rights groups calling for clearer transparency and explicit consent from creators whose work may be used to train AI systems.
English
