The use of copyrighted work as training data for artificial intelligence (AI) systems has invokes a legal debate regarding the boundaries of copyrights violation. By using doctrinal and comparative research methods, this article analyzes how this problem is handled in the United States of America (USA) through the case Getty Images v Stability AI and Vietnam to draw an example for Vietnam's developing Intellectual Property (IP) law. The research indicates that the USA issues surrounding training AI through copyrighted works are handled within the realm of fair use, which allows for flexible usage of copyrighted works, and fast adaptive use of copyright works, providing more technology and transformational uses. In contrast, Vietnam's new IP Laws are more reactive by attempting to provide AI training through the use of technology, but is vaguer in trying to provide actionable steps to technology and balancing within the system. By the fair use system and Vietnam’s closed model systems, this article contributes to ongoing law reform process by offering policy recommendations. Finally, this article aims to promote Vietnam’s developing IP framework, providing more certainty and balanced protection for copyright owners while fostering AI innovation.
Copyrights © 2026