The rise of AI-generated Ghibli styles reopens the debate on copyright and the trivialization of culture.
Marga Zambrana, UOC
The popularization of ChatGPT's image generator, capable of recreating recognizable artistic styles such as Studio Ghibli's, has sparked a new outbreak of debates in the legal and ethical fields. Beyond the viral phenomenon, the implications of this technology touch on fundamental issues of copyright, unfair competition, trivialization of art and cultural extractivism.
Some AI platforms are using protected works to train AI models without remunerating creators, while other businesses are trying to set up compensation mechanisms. The debate revolves around whether we are facing a case of misappropriation or a new way of generating shared wealth. Emerging regulations in the EU and the US could shape the future of this sector. Experts from the Open University of Catalonia (UOC) analyse the infringements arising from the massive use of recognisable visual styles from a legal point of view, and warn of cultural trivialisation and the loss of artistic identity in the face of aesthetic standardisation driven by AI.
Copying without copying: the legal challenge of protecting a style
The visual style of a work is not directly protected by copyright, neither in Spain nor in the European Union. However, when a set of aesthetic features is so identifiable that it acts as a "visual signature", it can enjoy indirect protections. "Even if it is not registered as a trademark, the law protects that identity to prevent other companies from taking advantage of its fame and confusing the public," explains Begoña González Otero, associate professor at the UOC's Faculty of Law and Political Science.
From a legal point of view, the massive use of Studio Ghibli's style by AI platforms opens the door to possible claims for unfair competition and appropriation of artistic identity. "Copying a style so associated with its author or study en masse can be illegal, even if a specific work is not copied," says González Otero. According to Article 6 of the Unfair Competition Law (LCD), inducing confusion or exploiting the reputation of others constitutes an illegal practice. There may also be a risk of dilution of the unregistered trademark if it is demonstrated that the style is distinctive and well-known.
The situation changes depending on who generates the images. "When there are platforms or companies that use recognizable styles to train or produce content that competes in the market, there is a clear risk of misuse," says the expert. However, if they are private users who create images for private use, "any direct action against them would be disproportionate", according to Directive 2004/48/EC.
In addition, the passage of the Digital Services Act (DSA) introduces new obligations for certain artificial intelligence platforms. Specifically, these rules apply when they act as intermediaries, i.e. if they allow users to generate, share or disseminate online content, such as images, texts or videos.
In these cases, the DSA – which extends the framework of the e-Commerce Directive – obliges platforms to act diligently on illegal content, to be more transparent and to better protect users' rights. If the platform exceeds 45 million users in the European Union, it is considered a very large online platform (VLOP) and must meet additional requirements, such as conducting systemic risk assessments (e.g. manipulation of information or mass dissemination of disinformation).
Now, if a platform doesn't just host third-party content, but uses that content to retrain its own models, it's no longer just a middleman. In that case, you assume direct liability, which can have stricter legal consequences.
Despite this regulatory framework, practical implementation remains limited. González Otero warns that, "although the legal bases exist, enforcement in Europe is weak, and the market favours opportunistic behaviour". Cases such as GEMA v. OpenAI illustrate the difficulty of defending copyright against AI models.
Cultural extractivism and the trivialization of art
From the artistic and ethical fields, the risks of this technology are not minor. Quelic Berga Carreras, a professor at the UOC's Faculty of Computer Science, Multimedia and Telecommunications and a researcher in the DARTS group, believes that the "ghiblification" of the internet represents a flagrant case of cultural appropriation. "We are asking the machine to copy the style without prejudice or filters, with which we trivialize an art that is based on care, detail and respect," reflects Berga.
The researcher warns that this phenomenon is not a simple aesthetic homage, but a cultural extractivism that trivializes profoundly human references. In his view, AI technology, as it is being used, reproduces a form of colonialism over alien cultures, by extracting recognizable traits and stripping them of their deep meaning for mass exploitation.
Three great risks for artistic creation
One of the main risks is the loss of creative self-esteem among artists. According to Berga, facing a machine capable of imitating styles in a spectacular way but without a soul can demotivate human creators and affect their confidence and motivation. Secondly, there is a progressive cultural blurring: the indiscriminate crossing of styles, fostered by AI, generates an "extreme hybridisation" that erases solid cultural references and turns creation into a superficial collage without roots.
Finally, Berga warns of a growing trivialization of art: the more spectacular images generated by AI are propagated, the more they are emptied of deep content, by replacing the social sense of art with the mere aesthetic appearance. Hayao Miyazaki himself, founder of Studio Ghibli, criticized in 2016 the use of AI to replicate human movements, exclaiming: "You don't understand the meaning of human effort." A criticism that resonates today more than ever.
AI and Copyright: A Race Against Time
European regulation is trying to adapt to the new scenario with measures such as the Artificial Intelligence Act (Regulation 2024/1689), which obliges providers of general-purpose AI models (GPAI) to respect copyright and detail the data used in the training of their models. However, Begoña González Otero warns that "there are ambiguities and gaps that can influence the effectiveness of obligations".
In fact, the lack of effective harmonisation between the Artificial Intelligence Act, Directive 2019/790 and national legislation generates a "perverse incentive": non-compliance is, for the time being, more profitable than complying.
For their part, artists and creators are also looking for strategies to protect themselves. Projects such as Tracking, detection and management of AI infringement (TDMAI) arise as technological initiatives promoted by the authors themselves to monitor the unauthorized use of their works.
Between fascination and risk
The fascination with the capabilities of artificial intelligence coexists with a growing ethical and cultural concern. Quelic Berga concludes that "the risk is not only to lose control over creation, but to get used to an aesthetic mediocrity that trivialises art and culture".
Meanwhile, the future of creativity and copyright in the age of AI remains open, in a race in which regulation, cultural awareness and collective action will be decisive in balancing innovation and respect for cultural identity, according to experts.
