OpenAI's Viral Studio Ghibli Image Generator Sparks Fresh Copyright Debates
A new ChatGPT image tool allows users to easily recreate copyrighted art styles like Studio Ghibli, pushing AI copyright questions back into the spotlight. Legal experts note that while mimicking an art style is not explicitly illegal, the underlying training methods remain a legal gray area.
OpenAI launches a new native image generator in ChatGPT that quickly goes viral as users flood social media with pictures transformed into the style of Studio Ghibli. People upload existing photos of celebrities, politicians, and movie scenes, prompting the AI to recreate them in the signature look of the beloved Japanese animation studio. This rapid adoption follows a similar trend sparked by Google's Gemini Flash model earlier in March.
The extreme ease of copying a distinct artistic style with a simple text prompt raises significant concerns about how these AI systems are developed. Intellectual property lawyer Evan Brown explains that while art style itself does not receive explicit protection under copyright law, the legality hinges on how OpenAI trains its models. It is highly plausible that OpenAI achieves this accurate likeness by ingesting millions of copyrighted frames from Ghibli films into its training databases.
This viral moment highlights the ongoing legal gray area surrounding generative AI and fair use. Courts are currently deciding whether scraping the web to copy copyrighted works for AI training constitutes infringement, a central issue in active lawsuits brought by The New York Times and other publishers against OpenAI. Until these landmark cases reach a conclusion, AI image generators continue to operate in a murky legal landscape.