Timelines go Ghibli-crazy as latest OpenAI tool sets Miyazaki vision loose on the world

OpenAI’s latest image generation tool sparked a viral trend of Studio Ghibli-style renderings, flooding social media with cartoons of pets, families, and famous landmarks or events reflecting the work of legendary auteur and animator Hayao Miyazaki -- only for the AI company to restrict access to Plus and Pro subscribers, within 24 hours.

By Wednesday night, social media timelines had transformed into vibrant galleries of whimsical imagery as users rushed to experiment with OpenAI’s newest image generation capabilities.

By Thursday, many users found their creative endeavors met with rejection messages. OpenAI had begun blocking requests from ChatGPT accounts which were still on the free tier for images in the “style of Ghibli” and some other artists — a swift response to concerns about creative rights that demonstrated the precarious balance between innovation and ethical responsibility in the rapidly evolving AI landscape.

The viral sensation highlighted a significant evolution for ChatGPT as it competes in the increasingly crowded AI chatbot market. Image generation isn’t unique to OpenAI’s offering — competitors such as xAI’s Grok and Google Gemini already feature similar capabilities — but the quality and precision of GPT-4o’s creations potentially set a benchmark that was evident by how popular it became.

“GPT-4o image generation excels at accurately rendering text, precisely following prompts, and leveraging 4o’s inherent knowledge base and chat context—including transforming uploaded images or using them as visual inspiration,” OpenAI said in its update documentation.

The tool’s distinctive strength lies in its granular understanding of complex prompts. While “other systems struggle with around 5-8 objects,” according to OpenAI, GPT-4o can handle requests featuring up to 20 different elements with remarkable accuracy. This precision stems from extensive human labelling of training data, creating what the company describes as a “tighter binding of objects to their traits and relations” that allows for unprecedented control.

To create these images, users simply upload a photo or describe a scene with text prompts such as “Visualise this photo into a Studio Ghibli-style anime illustration with soft textures, warm colours, and whimsical details.” Within seconds, GPT-4o generates an image that captures the essence of the Oscar-winning Miyazaki’s signature aesthetic — soft colour palettes, detailed natural settings, and lush backgrounds reminiscent of films such as Spirited Away, My Neighbour Totoro, Howl’s Moving Castle and The Wind Rises.

The swift Ghibli-style generations captured the tenuous relationship today between technological innovation and creative rights. That this emerged around Studio Ghibli’s aesthetic makes for a more striking capture of this challenge: the Japanese animation house occupies a unique position in global cinema — transcending both the animation genre and its cultural origins to achieve a rare universality, from mainstream culture to those with more nuanced and niche appreciation of the arts.

When Disney conquered Western animation with optimistic fairy tales, Miyazaki and his collaborators crafted narratives addressing environmental destruction, pacifism, and female agency without simplistic moralising. Each meticulously hand-drawn frame represents countless hours of human artistry — precisely the kind of creative labour that raises thorny questions about AI’s relationship to original artistic expression.

This isn’t the first time style transformation tools have captured public imagination. In 2016, the Prisma app gained rapid popularity for its ability to apply artistic filters inspired by painters like Pablo Picasso and Edvard Munch. The iOS version was downloaded 7.5 million times in its debut week, with the Android release achieving 1.7 million downloads on its first day — impressive figures in the pre-AI boom era.

However, OpenAI’s latest technology’s impressive realism also raised concerns about potential misuse. To address this, OpenAI confirms all image generations will adhere to C2PA (Coalition for Content Provenance and Authenticity) metadata guidelines, allowing viewers to distinguish between AI-generated and authentic images.

“What we’d like to aim for is that the tool doesn’t create offensive stuff unless you want it to, in which case within reason it does,” said Sam Altman, CEO of OpenAI. “As we talk about in our model spec, we think putting this intellectual freedom and control in the hands of users is the right thing to do, but we will observe how it goes and listen to society.”

OpenAI is particularly mindful of artistic rights following previous controversies and lawsuits. “We’re respecting of the artists’ rights in terms of how we do the output, and we have policies in place that prevent us from generating images that directly mimic any living artists’ work,” says Brad Lightcap, COO of OpenAI.

This commitment appears to have motivated the company’s swift response to the Ghibli-style trend. When questioned by news reporters about the restrictions, OpenAI explained that they had “added a refusal which triggers when a user attempts to generate an image in the style of a living artist.”

The latest ChatGPT update represents a significant transition from text-only or externally dependent image generation tools to fully integrated multimodal systems. It also reflects the accelerating pace of AI development, with companies racing to enhance their offerings. Google’s Imagen 3 model powers Gemini’s image generation capabilities across web and smartphone apps, while xAI’s Grok 3 has included image generation since early 2025.

OpenAI initially intended to make the new image generation capabilities available across all subscription tiers, but the overwhelming demand has forced a change of plans. “Roll-out to our free tier is unfortunately going to be delayed for a while,” Altman confirmed. For now, only ChatGPT Plus and ChatGPT Pro subscribers will maintain access to these features.