As generative artificial intelligence (AI) continues to take the spotlight, a group of Japanese experts have founded a think tank to prevent copyright issues arising from the use of AI models in the creative process.
The think tank, also called the ‘Anime Chain Preparatory Committee’, consists of AI and blockchain experts in Japan. In a report from Animation Magazine, the think tank builds an AI model with opt-in training data that anime creators can use without the hassle of copyright claims.
The model will rely on blockchain to ensure the “transparency and legitimacy” of training data, with processes in the chain recorded for clarity. Anime Chain says its main goal is to create a safe environment for creators, which it aims to achieve by seeking explicit permission from copyright holders.
“With our goal of respecting creators’ rights, we will provide a generative AI that is rebuilt from the base model and uses only opt-in materials that have received permission from the rights holders, along with related tools,” read the report
In addition to preventing intellectual property issues, the preparatory committee is also keen to solve the problem of labor shortage in the Japanese anime industry. Despite the rising population, there has been a steady decline in the number of animators in the industry, but the introduction of an AI system can keep the content mill turning.
The think tank says the new platform will return AI revenue to both creators and original IP rights holders, with an overview of all keyframes and images captured on a publicly distributed ledger.
Anime Chain has unveiled a three-step plan designed to help achieve its lofty goals, starting with establishing a collaborative framework for creators and industry players. The framework will guide the use of AI-generated content while creating revenue sharing rules.
The second phase of the plan involves the creation of a fundamental model trained with data obtained with permission from copyright holders. In the third phase, the think tank will use blockchain to onboard new creators and IP holders while pursuing a decentralized governance model.
Copyright issues threaten the future of generative AI
Across the Pacific, several creators have taken AI developers to court for using illegally sourced material to train their large language models (LLMs). OpenAI, Meta (NASDAQ: META) and Anthropic are defending themselves against injured intellectual property rights holders in US courts, with the industry keeping an eye on the outcome of the legal proceedings.
In 2023, Japanese experts warned that the lack of AI copyright rules could lead to an avalanche of legal battles in the near future. In June, Japanese MP Takashi Kii raised the alarm about the lack of regulations protecting copyright holders from AI abuse.