Experts are calling for Vietnam’s revised intellectual property law to include protection of artists’ voices, images, and creative styles against AI-generated deepfakes and unauthorized use by foreign platforms.

At a session discussing amendments to the Law on Intellectual Property (IP) on the morning of November 5, Associate Professor Dr. Bui Hoai Son, Standing Member of the National Assembly’s Committee on Culture and Education and a representative from Hanoi, stressed that the revised law must not only improve technical legal mechanisms but also act as a cultural firewall in the digital age.

Such reforms, he said, are crucial to protecting creators and asserting Vietnam’s cultural sovereignty in the global digital space.

Protecting artists’ digital identities from exploitation

W-Bùi Hoài Sơn Hà Nội 2310251.JPG.jpg
Associate Professor Dr. Bui Hoai Son - Standing Member of the National Assembly’s Committee on Culture and Education. Photo: Hoang Ha

Son supported the amendment of Clause 13, Article 4, which expands the scope of industrial design protection to include “non-physical products.” However, in a digital context, he argued, this must go beyond 3D models and also encompass digital images, voice profiles, performance styles, and digital personas.

He proposed explicitly adding such protections in the legal text to ensure that artists’ voices, likenesses, and creative styles are not copied, deepfaked, or commercially exploited by AI without consent.

“We must not allow Vietnamese artists to become free data suppliers for foreign AI companies,” Son emphasized, warning that without legal safeguards, cultural value may be reduced to a raw resource for global tech exploitation.

Referring to Clause 2, Article 7 of the draft, which prohibits IP rights from obstructing the use of national symbols like the flag, emblem, and anthem, Son recommended adding language to ensure this applies to both online environments and digital platforms. He also urged clarifying the responsibilities of individuals, organizations, and platform providers.

“This is how we safeguard the nation’s sacred symbols in the era of algorithms and AI-driven content moderation,” he said.

Regarding IP inspection, the revised Article 201 broadens the scope of subjects allowed to conduct assessments. But in an age where music, images, and content can be replicated by AI in seconds, Son argued that human evaluators alone are insufficient.

He proposed the inclusion of “digital inspection capabilities,” such as content recognition tools, AI data analysis, origin tracing, and copyright verification technologies. Without these tools, he warned, even the best laws would be powerless to resolve digital copyright disputes in practice.

He also called for the addition of emergency dispute resolution mechanisms for digital content and encouraged the use of technologies like content ID, watermarking, or blockchain to improve copyright enforcement.

Defining authorship and intent in the AI age

dbqh le hoang anh gia lai .jpeg
Representative Le Hoang Anh (Gia Lai delegation). Photo: Quang Khanh

Also speaking at the session, National Assembly member Le Hoang Anh (Gia Lai delegation) said that while the revised IP law includes positive developments, it risks falling behind in the rapidly evolving fields of data and artificial intelligence.

He highlighted the challenge of defining authorship and ownership in AI-generated works. Currently, the law recognizes authors as individuals who directly create content. However, in practice, many works today - ranging from music and art to journalism and analysis - are created entirely or predominantly by AI.

“If we continue using this traditional framework, Vietnam risks being left out of the digital creative wave because we won’t be able to determine rights and responsibilities when disputes arise,” he said.

He proposed adding a new clause defining “AI-generated works,” stating that only content created with significant human direction or intervention should qualify for protection. In this framework, the person who operates, programs, or controls the AI should be considered the legitimate author. Meanwhile, outputs fully generated by AI without human involvement should be treated as data or non-copyrightable innovations.

Referring to Clause 5, Article 7, which allows organizations and individuals to use public data to train and develop AI, Hoang Anh praised the inclusion but warned that it remains vague and fails to clearly distinguish between commercial and non-commercial use.

“If left unchecked, AI companies could mine public literary, musical, and journalistic databases created by millions of authors - without permission or compensation - then monetize those outputs,” he warned.

He recommended restricting public data use to non-commercial research purposes unless the user obtains consent or collective licensing agreements from copyright holders. In the case of disputes, data sources should be made transparent.

He also emphasized that allowing AI to utilize public data must align with other legal frameworks such as the Cybersecurity Law, Personal Data Protection Law, and Information Technology Law. As such, he proposed establishing a registration or disclosure mechanism for AI systems trained in Vietnam, following the models currently adopted by Australia and the European Union.

Thanh Hue