The day has only passed since the recent AI ChatgPT image generator began, and the channels in social media are already flooded with memes generated by AI in the style of studio Ghibli, the iconic Japanese studio of the animation standing behind films, such as “My neighbor Totoro” and “Spirited Away”.
Over the past 24 hours, we have seen images generated by AI representing the Ghibli Studio versions Elon Musk“Lord of the Rings“, And President Donald Trump. It seems that the general director of Opeli, Altman himself Has made his new profile photo studio image in the Ghibli stylePresumably made with a native GPT-4O image generator. It seems that users send existing images and photos to chatgpt and ask chatbot to recreate it in recent styles.
The latest OPENAI update appears on the heels of the release of a similar image function AI in its Flash Gemini model, which also caused a viral moment earlier in March, when people used it to remove water signs from images.
The latest Opeli and Google tools make it easier to restore styles of copyright protected by copying-by entering the text hints. Together, these recent AI image functions seem to revive the fears of several processes against generative programmers of AI models. If these training companies in the field of copyright protected work or do they violate copyright?
According to Evan Brown, an intellectual lawyer at the Neal & McDevitt law firm, products such as the native GPT-4O image generator operate today in legal gray. Style is not clearly protected by copyright, according to Brown, which means that Opeli does not seem to break the law simply generating images that look like ghibli studio films.
However, Brown claims that it is likely that Opeli has achieved this similarity by training its model on millions of frames from Ghibli movies. Even if this was the case, several courts still decide whether the training of AI models on protected work is protected by the permitted apply.
“I think it raises the same question that we have asked for several years,” said Brown in an interview. “What are the implications of copyright violations to go out, crawl with the Internet and copy to these databases?”
Recent York Times and several publishers are in energetic processes against Openai, claiming that the company has trained its AI models in the field of original work without proper assignment or payment. Similar claims were brought against other leading AI companies, including Meta and AI Strintup Midjourney.
In a statement for TechCrunch, Opeli spokesman says that although ChatgPT refuses to reproduce the “style of individual living artists”, OpenAi allows the replication of “wider studio styles”. Of course, it is worth noting that there are live artists who are assigned to the pioneer of the unique styles of their studio, such as co -founder Studio Ghibli, Hayao Miyazaki.
Apparently, users could also apply the native function of the GPT-4O image generation to recreate the styles of other studies and artists. Someone else made a portrait of Marc Andreessen in a style Dr. Seussand the marriage recreated them Pixar wedding photos.
We tested several popular AI paintings – including those available at Google’s Gemini, Xai’s Grok and Playground.Ai – to see their ability to match the Ghibli style studio style. We found that the recent Openai image generator created the most precise replica of the style of the animation studio.
For now, the recent features of OPENAI and Google image show a leap forward in what AI models can generate, which seems to boost apply. OpenAI delayed the implementation of a recent tool to the image to free level users on Wednesday, citing high demand. This may be the most significant for these companies today, but we will have to wait for the courts to burden their legality.