In this piece for TechCrunch, author Maxwell Zeff looks into the latest wave of AI-generated images flooding social media, just a day after ChatGPT’s new image-generation tool became available.
Since its launch, users have been filling their feeds with AI-crafted visuals in the signature style of Studio Ghibli, the renowned Japanese animation studio behind classics like My Neighbor Totoro and Spirited Away.
Within 24 hours, Studio Ghibli-inspired AI depictions of Elon Musk, The Lord of the Rings characters, and former President Donald Trump have gone viral. Even OpenAI CEO Sam Altman appears to have joined in, reportedly setting his profile picture to a Ghibli-style illustration—likely created using GPT-4o’s built-in image generator. Many users are experimenting by uploading existing photos and asking ChatGPT to reinterpret them in different artistic styles.
OpenAI’s latest feature arrives shortly after Google introduced a similar AI image tool in its Gemini Flash model, which made headlines in March when users discovered it could remove watermarks from images.
Both OpenAI and Google’s advancements highlight how effortlessly these tools can replicate well-known artistic styles with just a text prompt. However, a major controversy remains: Are these AI models trained on copyrighted material without permission? If so, does that constitute copyright infringement?
These concerns are central to several ongoing lawsuits challenging the legality of generative AI models and their use of copyrighted works in training data…