By: Atty. Hera Marie Ydulzura
When the world was forced into isolation during the pandemic, the internet became THE lifeline. It was one of the things that kept everyone connected. The toll of having all the time in the world with no one to physically share it with enabled individuals to explore or revisit their artistic sides. Sharing creations online became a source of joy and connection, with newbie and veteran artists praising each other for their creativity and helping each other grow artistically.
However, when the world opened its doors again after the pandemic, another door also swung wide open: the door to the world of generative AI.
Generative AI refers to machine-learning models or tools that can generate text, images, and other content based on the data they were trained on. Generative AI systems are distinguished from other AI systems by their ability to create novel output based on users’ prompts. Generative AI has rapidly evolved into a tool capable of composing music, writing stories, and producing artwork that can rival—if not, surpass—the creations of human artists.
And this is where the issue lies. Generative AI can only create based on the data they were trained on. It learns from massive datasets made from existing works from artists all over the world. Often, these works are used without the creators’ knowledge, consent, or compensation.
As a result, aside from concerns over data privacy, questions of copyright and ownership have been thrown around, yet no one has pinned down a clear answer to these concerns.
Take for example, the “Ghiblification” trend. Earlier this year, OpenAI’s new deep learning model jump-started the era of AI-generated art. The most viral of them all was the Ghiblification trend, which takes ordinary pictures and transforms them into the style of Studio Ghibli’s iconic animation style. Like Studio Ghibli, many artists around the world have expressed concerns over their unique styles being mimicked at the click of a button.
At the heart of this issue is copyright which is a form of legal protection given to the creators of original works such as art, music, films, books, and software. It gives the creator the exclusive right to decide how their work is used, reproduced, distributed, performed, or adapted. But with the advent of AI, the question remains: who owns ownership over these AI-generated works when it is built on the unpermitted use of countless human artistic creations?
In 2023, the Beijing Internet Court (BIC) laid down a landmark case (Li Yunkai v. Liu Yuanchun) by declaring that AI-generated image is copyrightable. This was affirmed by the Changshu People’s Court in 2025 (Lin Chen v. Hangzhou Gaosi Membrane Technology).
These two cases hold similar factual circumstances wherein the plaintiff used an AI tool to generate a picture by inputting prompts and adjusting the parameters. They then posted this on social media, where the defendant used and appropriated the picture as their own, without permission or compensation to the plaintiff.
In the 2023 case, the BIC discussed the concepts of “originality” and “intellectual achievement”, among other things, as requirements before a work can become copyrightable under the Chinese Law. “Originality” requires the resulting work to be independently completed by the author and reflect that author’s personalized expression. On the other hand, “intellectual achievement” is “the result of intellectual activities, so the work should reflect the intellectual input of a natural person.” In addition, in the 2025 case, the user agreement of the AI tool was likewise taken into consideration.
The Chinese courts concluded that the Plaintiff’s AI-generated work was entitled to copyright protection because the Plaintiff could guide and influence the resulting image to a certain extent, and by modifying the prompts entered into the AI tool, they provided their own human creative expression, which was then reflected in the final AI-generated image.
Interestingly, in a similar case of using generative AI for one’s work, the United States Copyright Office (USCO) reached a different conclusion.
In the case of “Zarya of the Dawn”, Kashtanova used an AI tool to produce the images for her graphics novel. She applied for and obtained copyright registration for her graphics novel without disclosing that she used artificial intelligence to create any part of the work nor any disclaimer to any portion of said work. However, when USCO learned through social media that she had created the graphics in the novel using an AI tool without disclosing such fact in her application, USCO expressed its intent to cancel the registration and gave Kashtanova a chance to explain.
In 2023, the USCO ultimately gave Kashtanova limited registration of her graphic novel. It held that although Kashtanova is the author of the work’s text as well as the selection, coordination, and arrangement of the work’s written and visual elements, the images in the work that were generated by the AI tool are “not the product of human authorship”.
In reaching such conclusion, the USCO reviewed how the AI tool processes and generates images and eventually determined that there is a significant distance between what the user may direct the AI tool to create and the resulting image that the AI tool produces. The USCO likened this to commissioning a visual artist—you can give general directions and/or detail how you want the output to appear, but the visual artist will still determine how to best express them, and as such is the author of that image. Thus, contrary to the Chinese court ruling, the USCO held that AI-generated images are not copyrightable.
But what about in the Philippine legal landscape? Considering that the Philippines is known as the “social media capital of the world”, anyone with a smartphone and internet connection can create AI-generated art. No doubt some people took advantage of this and tried to sell such AI-generated art on a “commission” basis. Do these people really have the right over these AI-generated art (and subsequently sell them) even if all they did was click a few buttons here and there?
During a Big Bad Wolf event last 2023 in Manila, the Intellectual Property Office of the Philippines (IPOPHL) participated and had the opportunity to address questions regarding generative AI in one of its seminars. IPOPHL explained that under the current copyright laws, AI-generated works are not copyright-protected since only natural persons can be considered authors. Further, copyright in works that are partially AI-generated protects only those parts that are created by natural person. As such, IPOPHL relies on the declarations of the creators in their applications for copyright.
For reference, under the Intellectual Property (IP) Code of the Philippines, copyright protects original intellectual creations of an author who is a natural person including literary, artistic, dramatical and musical works, as well as computer programs and cinematographic and audiovisual works, and other works in the literary and artistic domain, from the moment of its creation. Although the Philippines does not yet have specific AI regulations on copyright, IPOPHL is working on guidelines in relation to AI-generated works, copyright protection, and copyright compliance.
In March 2024, the IPOPHL issued a new Fair Use Guidelines, clarifying statutory exemptions and the limits of fair use. For reference, the general principle of fair use, under Section 185 of the IP Code, permits the use of copyrighted works without prior permission from the copyright holder, depending on certain factors.
The new Fair Use Guidelines, on the other hand, tackle Section 184 as a general limitation on copyright to provide some clarity on conditions that must be met in order for specific acts to be considered as statutory fair use or in a manner that does not infringe on IP rights.
However, it does not appear that AI-generated work has already been addressed in this new guideline.
Nevertheless, the IPOPHL, through the former Director General Barba, expressed at the Philippine International Copyright Summit held last year, that it is planning to release guidelines that will ensure copyright protections in the application of artificial intelligence in creative works.
Additionally, it appears that IPOPHL will be taking a stance similar to that of USCO. In the said copyright summit, IPOPHL Bureau of Copyright and Related Rights Director Cuyo mentioned that “the focus of the guidelines will be on copyright registration, similar to what was released by the US Copyright Office sometime in 2010.”
“With questions on ethics and law, such as whether AI can be protected with copyright, whether AI works can be original as a derivative of existing copyrighted works, and whether works made with the help of AI should give credit to the software, its developers, and trainers, IPOPHL attempts for answers,” Mr. Barba said.
The debate over copyright in the world of generative AI remains unsettled. While courts in China have recognized certain AI-generated works as copyrightable, the United States has taken the opposite view, both based on whether or not human intellectual creation is attendant in their creation. Meanwhile, the Philippines is still in the process of deciding for itself on how to approach the issue, but is it highly aware that there is a growing need to safeguard both the rights of the creators and the innovators (of such AI).
While some embrace generative AI as a tool to support and expand their creative processes, others view it as a threat, particularly for those who rely on their art for their livelihood, undermining years of practice and personal expression. The door to the world of generative AI is a door that only swings one way. Whether that leads to development or exploitation of human creativity depends on how the government chooses to balance progress or the preservation of human creativity.