Every painting, musical composition, or film created by artificial intelligence (AI) is built upon a fundamental process of learning from vast repositories of human-generated data, commonly known as Big Data. Recent legal developments around the world, including in Vietnam, indicate that amid the accelerating technological race, a new principle is taking shape: the rights of creators must remain at the center of AI development and deployment.
Creative data cannot be presumed available for use
In mid-June 2026, the Motion Picture Association Asia Pacific (MPA Asia Pacific) sent a document to the Vietnam Association for Copyright Protection of Film and TV Movies (VAFC), requesting the public announcement that its member studios reserve their rights regarding the use of copyrighted works in AI-related activities. Behind the document lies a clear message: creative content is not a resource without ownership.
If, a few years ago, debates mainly centered on whether a painting or a song generated by AI could qualify for copyright protection, the focus has now shifted to the origin of the input data. This is because, to create any output, AI systems must first “learn” from vast datasets, or big data, that have been built and curated by human creators over many years.
The most important question today is whose data AI has used and whether the owners of the original data have given their consent, Le Thi Minh Hang, Director of the Center for Law and Copyright under the Vietnam Reproduction Rights Organization (VIETRRO), said.
This principle is also reflected in Decree No. 134/2026/ND-CP, which amends and supplements Decree No. 17/2023/ND-CP, guiding the implementation of the Law on Intellectual Property, and took effect on April 1, 2026. For the first time, Vietnamese legislation includes a dedicated section that more clearly defines the role of human beings in AI-assisted creative processes.
Under the decree, AI is regarded solely as a tool to support creativity. To qualify for copyright protection, an AI-generated work must demonstrate meaningful human creative involvement through the selection, direction, evaluation, editing, and final decision-making processes. In other words, AI may assist creativity, but it cannot replace the creative role of human beings.
The decree also establishes stricter requirements for the mining and use of texts and data for research, testing, and AI training purposes. These include ensuring that data is obtained through lawful access and that such use does not interfere with the normal exploitation of copyrighted works. Notably, copyright holders retain the right to reserve their rights and refuse access to copyright-protected materials for AI-related purposes.
Innovation and copyright protection must go hand in hand
Globally, approaches to preventing AI systems from freely accessing copyright-protected data sources continue to vary. Since 2019, the European Union has established a framework governing text and data mining within the digital market. Japan has adopted a more open approach, with its legal framework allowing AI systems, to a certain extent, to use data for information analysis and AI training purposes.
The United States has yet to enact a federal law specifically regulating data mining for AI training. As a result, disputes are generally resolved through court proceedings or private licensing negotiations between AI companies and rights holders. China, meanwhile, requires AI developers to ensure that training data is obtained through lawful channels and that intellectual property rights are fully respected.
While regulatory approaches differ, one point of convergence is becoming increasingly clear: AI cannot continue to advance by disregarding the rights of creators. As data emerges as a critical resource in the digital economy, protecting creators’ rights to choose how their works are used is not intended to hinder innovation but rather to ensure that innovation develops on a fair and sustainable foundation.
Behind every dataset that AI systems absorb lies the knowledge, labor, and emotions of individuals who have spent years creating cultural value. Respecting their right to make decisions about the use of their works is also a way of safeguarding creative motivation—the most important resource of any cultural industry in the AI era.
According to Mr. Luong Luu Bien, a prominent contemporary visual artist in Vietnam and a member of the Ho Chi Minh City Fine Arts Association, the value of art does not lie solely in the final product but in the creative journey itself, a process in which artists continuously make choices, experiment, make mistakes, revise, and refine their work. These are dimensions of creativity that data and algorithms can hardly replicate in their entirety.
