The viral Ghibli trend has taken the internet by storm, with millions of users transforming their personal photos into Ghibli-style images and sharing them across social media. While many are excited about the creative possibilities, serious concerns are being raised over the copyright implications of AI-generated art. Despite Studio Ghibli’s celebrated 40-year legacy, the rapid production of similar images has ignited a debate: could OpenAI or individual users be held liable under current copyright laws? In an exclusive interview with Tech Achieve Media, Sonam Chandwani, Managing Partner of KS Legal and Associates, breaks down these complex legal questions arising around the Ghibli trend and offers insights into what is protected by copyright law.
TAM: The Ghibli trend is currently doing the rounds on social media. Does AI-generated art that mimics a specific artistic style, such as Ghibli, constitute copyright infringement under current intellectual property laws?
Sonam Chandwani: Under the Copyright Act, 1957, copyright subsists in original artistic works that involve skill, labour, and judgment. The Act protects the expression of an idea, not the idea or the style itself. Therefore, a visual style, no matter how distinctive or recognisable, does not enjoy independent copyright protection under Indian law. However, if an AI-generated image reproduces or substantially resembles specific protected elements of a copyrighted work, such as unique character designs, backgrounds, or visual compositions, it may amount to infringement under Section 51 of the Act. The assessment will depend on whether the similarities are substantial in nature and whether the original expression is copied in a material manner.
Also read: The Dark Side of ChatGPT’s Ghibli Trend – Why It’s Not Good News
TAM: Since AI models are trained on vast datasets that may include copyrighted works, could companies like OpenAI be held legally accountable for generating images that resemble protected art styles?
Sonam Chandwani: There is currently no express provision in the Copyright Act, 1957, that directly governs liability arising from the training of artificial intelligence models on potentially copyrighted datasets. However, under Section 51, if a party facilitates or authorises the use of copyrighted works without permission and such facilitation results in public communication or reproduction, the company may be subject to claims of indirect infringement depending on the facts. While Indian law does not yet recognise contributory infringement in express terms, courts may evolve the doctrine in cases where there is clear knowledge and commercial intent. In the absence of statutory clarity, such matters would be subject to judicial discretion and evidentiary evaluation.
TAM: What legal recourse do studios like Studio Ghibli or individual artists have if their distinctive style is replicated by AI tools without permission?
Sonam Chandwani: If the AI-generated image results in copying of protected elements of an original work, the copyright holder may initiate a civil suit under Section 55 of the Copyright Act, 1957, seeking injunctions, damages, and account of profits. Criminal remedies may also be available under Section 63 if the infringement is proven to be wilful and done for profit. In addition to these remedies, artists may assert their moral rights under Section 57 of the Act, which protects against distortion, mutilation, or modification of their work in a manner that prejudices their honour or reputation. However, if only the style and not any identifiable expression is replicated, legal recourse may be limited under the current legal framework.
TAM: Do you foresee copyright laws evolving to better address AI-generated content, and what kind of regulations could be implemented to protect original creators?
Sonam Chandwani: It is likely that Indian copyright law will undergo amendments to address the growing influence of artificial intelligence in content creation. One potential area of reform could include defining authorship and ownership of AI-generated works under Section 2(d), which presently recognises only human authorship. Legislative changes may also introduce transparency requirements for datasets used to train artificial intelligence models and provide original creators with a mechanism to exclude their works from such datasets. Additional protections may also include a limited registration process for distinctive visual identities or stylistic signatures, thereby creating a new category of rights adjacent to copyright and design law.
TAM: Beyond copyright concerns, could AI-generated images in the style of Ghibli trend lead to trademark or moral rights violations, and how might courts approach these cases?
Sonam Chandwani: Under the Trade Marks Act, 1999, legal action may be pursued if the AI-generated image includes protected brand identifiers such as names, logos, or visual elements that are associated with a studio and registered as trademarks. Use of such identifiers in a manner that causes confusion or implies false association may constitute infringement under Section 29. Furthermore, moral rights under Section 57 of the Copyright Act, 1957, allow authors to object to any distortion, mutilation, or modification of their work that prejudices their honour or reputation. If artificial intelligence generates content that misrepresents the tone, integrity, or intention of the original artistic work, the author may invoke these moral rights. In determining such cases, courts in India are likely to examine whether the AI-generated content causes reputational harm and whether there is any unjust enrichment or consumer confusion arising from the unauthorised use of the original identity or elements.
