Japanese Publishers Demand OpenAI Stop Unauthorized Use of Copyrighted Works
A group of leading Japanese publishers, including the iconic Studio Ghibli, has officially called on OpenAI to halt the use of their copyrighted content in AI training without obtaining proper permission. This demand underscores increasing unease about how creative materials are being utilized in artificial intelligence growth without regulation.
Studio Ghibli’s Unique Animation Style Under Threat
Globally celebrated for films such as Princess Mononoke and The Wind Rises, Studio ghibli’s distinctive visual style has become a popular target for replication by OpenAI’s generative AI technologies. Since the introduction of ChatGPT’s integrated image generation feature in early 2025,users have frequently requested transformations of personal images-ranging from selfies to pet portraits-into artwork that mimics Ghibli’s signature aesthetic. Even OpenAI CEO Sam Altman joined this trend by updating his profile picture with a “Ghiblified” version.
The Rise of Video Generation and Sora App Sparks New Legal Worries
The launch of OpenAI’s Sora app, which offers video creation tools gaining popularity across Japan and internationally, has intensified concerns among rights holders. The Content Overseas Distribution Association (CODA) has publicly demanded that its members’ works not be exploited for machine learning purposes without explicit consent. CODA warns that using copyrighted content during AI training could violate Japan’s intellectual property laws.
Navigating Legal Uncertainties Around AI and Copyrighted Material
OpenAI often adopts a “seek forgiveness rather than permission” approach when incorporating copyrighted works into its training datasets. This strategy has enabled widespread production-and sometimes misuse-of images featuring protected characters or even deceased public figures through deepfake functionalities embedded within apps like Sora.
This practice faces pushback from major corporations such as Nintendo and estates managing legacies like Dr. Martin Luther King Jr.,who are concerned about unauthorized digital impersonations facilitated by these platforms.
Diverse global Approaches to Copyright Enforcement in AI contexts
The legal landscape surrounding copyright use in artificial intelligence remains unsettled worldwide. In the United States, where copyright law largely stems from statutes enacted in 1976, courts continue to wrestle with applying existing rules to machine learning data sets. A recent federal decision found Anthropic did not infringe copyrights by training on protected books but was penalized for illegally acquiring copies during data collection-a subtle yet significant distinction illustrating ongoing complexities.
In contrast,Japan’s CODA maintains that reproducing specific copyrighted elements within outputs generated by systems like Sora 2 likely breaches domestic intellectual property regulations due to stringent requirements mandating prior authorization before usage.
Miyazaki’s Longstanding Critique Reflects Wider Industry Concerns
Hayao Miyazaki-the visionary director behind many beloved Studio Ghibli classics-has long voiced strong disapproval toward certain computer-generated animations resembling his work. Nearly ten years ago, upon encountering early examples of AI-driven 3D animation inspired by his style, he condemned it as “utterly repugnant” and an insult “to life itself,” highlighting deep-seated fears about automated reinterpretations undermining artistic authenticity.
Toward Ethical innovation: Balancing Creativity with Intellectual Property Rights
- The swift rise of generative AI demands clearer frameworks that respect creators’ rights while encouraging technological advancement.
- CODA insists companies like OpenAI secure explicit permissions before integrating proprietary content into their model training processes.
- This situation exemplifies broader global challenges related to protecting intellectual property amid rapid progress in artificial intelligence capabilities.
- An estimated 47% surge occurred globally between 2024-2026 in reported incidents involving unauthorized digital reproductions linked directly to emerging generative AI technologies.*
- *This figure is based on aggregated reports from international IP enforcement bodies monitoring trends associated with new applications powered by generative AI.*
