Tiếng Việt
中文 (中国)
日本語
The rapid development of artificial intelligence (AI), particularly generative AI models, is bringing profound changes to the music industry. Whereas composing a song previously required advanced professional skills and an extensive creative process, AI now enables users to generate a complete song within a very short period of time. AI is no longer limited to suggesting creative ideas but can participate in the entire music production process, from composing melodies and writing lyrics to arranging music and even simulating a singer’s voice.
In practice, the use of AI to create music is becoming increasingly widespread in various forms. Products such as “AI music,” “AI covers,” and songs featuring AI-generated imitations of famous singers’ voices are appearing with increasing frequency on digital platforms such as YouTube, TikTok, and online music streaming services. These products are not only created for personal entertainment but are also commercially exploited through online distribution, performances at enterprises doing business in Vietnam, and monetization on digital platforms.
However, this rapid development has raised an important legal question: Does using AI to create music constitute infringement of copyright in Vietnam or infringement of related rights? In some cases, AI merely serves as a supporting tool that helps humans realize their creative ideas. In many other cases, however, AI can reproduce melodies, artistic styles, or even an artist’s voice without the authorization of the rights holder. This has raised significant concerns regarding intellectual property infringement, particularly when the resulting works are commercially exploited.
The issue becomes even more complex because current legislation, including Vietnam’s Intellectual Property Law, does not contain specific provisions directly governing music generated by AI. Traditional concepts such as “author,” “work,” and “creation” were developed based on human creative activity and are now being challenged by the increasingly significant role of AI. Consequently, determining the boundary between lawful conduct and infringement in the context of AI-generated music has become an urgent issue, not only for lawmakers but also for individuals and businesses utilizing this technology.
Against this backdrop, examining and clarifying the question, “Using AI to Create Music and Songs: Copyright Infringement or Lawful Innovation?” is of considerable importance in shaping an appropriate legal approach while maintaining a balance between intellectual property protection and the promotion of innovation within the music industry.
Overview of AI Applications in Music Creation
As AI technology continues to advance, its application in music creation is no longer confined to a single approach but has evolved into a wide variety of models, reflecting different levels of collaboration between humans and machines.
The first model involves fully AI-generated compositions. Under this approach, users simply provide basic prompts, such as a theme, mood, or musical style, after which the AI system automatically generates the entire musical work, including the melody, lyrics, harmony, arrangement, and even vocal performance. Human involvement in this scenario is minimal, generally limited to initiating the request or selecting the preferred output. This form of AI usage has generated the greatest legal controversy because it contains little or no direct creative contribution from a human author.
The second model involves AI-assisted creativity. This is the more common approach in practice, where AI functions as a tool supporting musicians and producers throughout the creative process. Users may employ AI to suggest melodies, refine lyrics, arrange musical compositions, or enhance recordings. In these circumstances, the final work still reflects substantial human creativity, while AI serves primarily as a technical tool, similar to traditional music production software but with a significantly higher level of automation.
Another rapidly developing application is AI-based simulation and reproduction. This technology enables the creation of “AI covers,” in which the voice of a particular singer is simulated to perform an entirely different song. AI is also capable of reproducing an artist’s musical style, melodic techniques, or arrangement structure. This area is particularly sensitive because it directly concerns an artist’s personal identity and commercial value while also creating significant risks of infringing related rights.
It should also be noted that most AI music systems currently operate by training on existing datasets. These datasets typically include sound recordings, songs, vocal performances, and various musical elements, many of which may be protected by Vietnam copyright registration. The use of such materials, particularly without authorization from the rights holders, raises significant legal concerns regarding unauthorized reproduction or exploitation of protected works.
Finally, the intended purpose of use is another important factor requiring distinction. Creating AI-generated music for personal use, research, or entertainment may be viewed differently from commercial exploitation through publication, public performance, or monetization on digital platforms. The wide variety of AI applications and intended uses has made assessing the legality of AI-generated music increasingly complex, requiring a case-by-case legal analysis.
Legal Framework Governing Copyright and Related Rights
Determining whether AI-generated music constitutes lawful use or copyright infringement requires analysis within the framework of existing copyright and related rights legislation.
First, copyright protection is granted to works directly created by human authors and expressed in a tangible form. One of the fundamental requirements for copyright protection is that the work must be original and reflect the author’s individual creative expression. Consequently, under current law, works generated entirely by machines or algorithms without significant human creative contribution are generally not recognized as copyrightable subject matter.
In addition to copyright, related rights play an equally important role in the music industry. Related rights include performers’ rights over their performances, as well as producers’ rights over sound recordings and audiovisual recordings. These rights exist independently of copyright and may arise even where the underlying work is not protected or where its Vietnam copyright registration has expired.
A fundamental principle of intellectual property law is that protection extends only to the specific expression of a work rather than to ideas, artistic styles, or creative trends. Accordingly, if AI reproduces the musical style of an artist without directly copying specific protected content from an existing work, such conduct may not constitute copyright infringement under the traditional interpretation of copyright law. However, this principle also exposes a significant legal gap, particularly as AI technology becomes increasingly capable of sophisticated stylistic imitation.
Furthermore, intellectual property law clearly provides that acts such as reproducing, publicly performing, communicating works to the public, or commercially exploiting copyrighted works generally require the authorization of the author or copyright owner, except where statutory limitations or exceptions apply. This principle is particularly relevant when evaluating AI-generated musical works that incorporate or are based upon copyrighted materials.
One of the greatest legal challenges today, however, is that existing legislation does not specifically regulate the use of copyrighted works as training data for AI systems. Whether such use constitutes reproduction or unauthorized exploitation remains an unresolved legal issue, not only in Vietnam but also across many jurisdictions worldwide.
These principles demonstrate that although the current legal framework governing copyright and related rights is relatively comprehensive with respect to traditional forms of creative expression, the emergence of AI presents a range of entirely new legal challenges. Addressing these challenges will require either flexible legal interpretation or legislative reform to ensure that intellectual property law remains compatible with rapidly evolving technological developments.
Foreign investors conducting business in Vietnam should consider consulting ASL LAW for tailored Legal Advice on Business oversea that addresses their specific commercial and operational requirements.
When Is the Use of AI Lawful?
The use of AI to create music is not inherently unlawful in every circumstance. In practice, if certain conditions are satisfied, AI-generated music may be considered lawful under the framework of copyright and related rights law.
First, AI should function as a creative assistance tool rather than a complete substitute for human creativity. Where the user actively participates in the creative process—such as developing the concept, refining melodies, editing lyrics, or shaping the musical style—the final work retains the creator’s personal creative contribution. In such cases, the individual may be recognized as the author and enjoy copyright protection over the resulting work.
Second, the data used to train AI systems should be lawfully obtained and used. If the input data consists of materials in the public domain, works licensed for such use, or works whose copyright protection has expired, the use of AI to generate new musical content is generally less likely to infringe the rights of third parties. Conversely, the unauthorized use of copyrighted sound recordings or musical works for AI training may give rise to legal liability.
Furthermore, AI-generated music should not unlawfully reproduce or create unauthorized derivative works based on existing copyrighted material. If the melody, lyrics, or musical structure bears substantial similarity to a protected work, the resulting AI-generated content may still constitute copyright infringement, regardless of the fact that it was created using AI technology.
Another essential consideration is the protection of related rights, particularly the rights of performers and producers of sound recordings. The unauthorized use of protected vocal performances, sound recordings, or specific audio elements may result in legal liability, even where the final product has been generated through AI.
Finally, the intended purpose of use also affects the legal assessment. In certain circumstances, AI-generated music created for personal, research, or other non-commercial purposes may be treated more flexibly than music created for commercial exploitation. Nevertheless, even non-commercial use may still constitute infringement if it violates copyright or related rights.
When Can the Use of AI Infringe Copyright or Related Rights?
Alongside lawful applications, the use of AI in music production also presents significant risks of infringing copyright and related rights if not properly managed.
One of the most common concerns is the use of copyrighted materials to train AI systems without authorization from the relevant rights holders. AI training generally involves copying and analyzing sound recordings and musical works, which may constitute unauthorized reproduction unless supported by an applicable legal basis.
In addition, generating AI-created music that constitutes an unauthorized derivative of an existing work presents another significant legal risk. Where AI produces melodies, lyrics, or arrangements that are substantially similar to a particular copyrighted work, the resulting output may be regarded as unauthorized reproduction or adaptation, even in the absence of direct copying.
Another highly debated issue concerns AI-generated vocal imitation. If AI systems rely on original sound recordings to train or recreate a singer’s voice, such activities may infringe the related rights of performers and producers of sound recordings. Even where original recordings are not directly used, AI-generated voice cloning may still affect an artist’s reputation and commercial interests, although current legislation does not yet provide a comprehensive legal framework governing such situations.
Similarly, the use of AI-generated sound recordings without the authorization of the relevant producers may also infringe related rights, particularly where such recordings are commercially exploited through distribution, public performance, or monetization.
A common misconception is that “AI-generated music belongs to no one” and can therefore be freely used. In reality, this assumption is not entirely accurate. Many AI-generated works still involve human creative input or incorporate elements protected by other intellectual property rights, meaning unauthorized use may still expose users to legal liability.
Current Legal Grey Areas
Although certain legal principles already exist, their application to AI-generated music remains uncertain in several important respects.
The first grey area concerns music generated entirely by AI without meaningful human creative input. Since copyright law generally requires human authorship, such works are typically not eligible for Vietnam copyright registration. However, this raises further questions regarding who, if anyone, may lawfully exploit or control these AI-generated outputs and whether any related rights may arise.
A second grey area involves collaborative creation between humans and AI. Where a musical work results from a combination of human creativity and AI-generated content, determining whether the human contribution is sufficiently original to justify copyright protection can be difficult, potentially giving rise to ownership disputes.
Another unresolved issue concerns AI-generated voice imitation. Current legal frameworks generally do not recognize a person’s voice as an independently protected intellectual property asset, making it difficult to address AI voice cloning, particularly where no direct copying of original recordings has occurred.
Likewise, AI-generated imitation of an artist’s musical style is not easily classified as copyright infringement because copyright law does not protect artistic style itself. Nevertheless, as AI systems become increasingly capable of accurately replicating the distinctive musical characteristics of individual artists, the boundary between lawful inspiration and unlawful appropriation becomes increasingly blurred.
In addition, the legality of using copyrighted materials to train AI systems remains largely unresolved. This represents one of the most significant legal challenges today, as determining whether AI training constitutes lawful use or copyright infringement often depends on the interpretation of existing legal principles in individual cases.
Finally, proving infringement in the AI environment presents substantial evidentiary challenges. AI-generated outputs may be substantially similar in essence to protected works without involving direct copying, making it considerably more difficult for rights holders to establish infringement and effectively enforce their rights.
Practical Implementation and Challenges in Vietnam
In Vietnam, the development of AI-generated music has accelerated rapidly, particularly on digital platforms such as YouTube and TikTok. Music created entirely or partially with AI assistance is being distributed at an unprecedented pace, attracting significant public attention and encouraging wider adoption across both creative and commercial sectors.
However, this rapid dissemination has also created considerable challenges for legal enforcement. In many cases, it is difficult to determine who should be recognized as the author, who owns the intellectual property rights, or who should bear legal responsibility for AI-generated musical content. Such uncertainty reduces the effectiveness of existing mechanisms for protecting copyright and related rights.
At the same time, the role of digital platforms remains relatively limited. Although current legislation imposes certain obligations on intermediary service providers, the implementation framework is not sufficiently detailed, particularly with respect to identifying and removing infringing content. The absence of an effective “notice-and-takedown” mechanism allows many infringing works to remain accessible and continue spreading across online platforms.
Another important issue concerns public awareness. Many individuals and organizations still have an incomplete understanding of the legal rules governing AI-generated music, resulting in the widespread use of AI tools without adequate consideration of potential copyright and related rights implications. This lack of awareness increases the likelihood of inadvertent legal violations.
Directions for Legal Reform and Practical Solutions
Addressing these issues requires both legislative improvements and enforcement mechanisms that reflect the realities of emerging AI technologies.
First, legislation should provide greater clarity regarding the legal concepts of authorship and copyright ownership where AI is involved in the creative process. Establishing clear standards for assessing the extent of human creative contribution would reduce legal uncertainty, minimize disputes, and promote greater consistency in the application of copyright law.
In addition, lawmakers should consider expanding legal protection for artists’ voices and distinctive performance characteristics. Such protection would provide stronger safeguards against unauthorized AI-generated voice cloning and other forms of digital imitation that may adversely affect performers’ commercial interests and professional reputations.
Another key area requiring attention is the regulation of AI training data. The legal framework should require greater transparency regarding the sources of training datasets and establish clear licensing mechanisms for the use of copyrighted materials. Such measures would help protect the interests of rights holders from the earliest stages of AI system development.
Furthermore, enforcement mechanisms should be strengthened, particularly through the establishment of an effective notice-and-takedown procedure for digital platforms. Clearly defining the responsibilities of online platforms would significantly improve the detection, removal, and prevention of infringing content.
At the same time, technological solutions should be incorporated into enforcement efforts. The use of audio recognition technologies, content identification systems, and advanced data analytics can substantially enhance the ability of authorities and rights holders to detect and address copyright infringements involving AI-generated music.
Finally, greater efforts should be made to improve public awareness of the legal framework governing AI and music creation. Educating creators, enterprises doing business in Vietnam, and consumers about applicable intellectual property rules will help reduce violations and encourage the responsible use of AI technologies.
Conclusion
The use of AI in music creation presents significant opportunities for innovation while simultaneously introducing complex legal challenges. As AI technology becomes increasingly capable of replacing or replicating creative elements traditionally associated with human authorship, conventional concepts of copyright and related rights are facing unprecedented pressure.
The foregoing analysis demonstrates that the use of AI is neither inherently lawful nor inherently unlawful. Its legality depends on a range of factors, including how AI is used, the legality of the training data, the extent of human creative contribution, and the intended use of the resulting musical work.
Looking ahead, continued legal reform and stronger enforcement mechanisms will play a critical role in establishing a clear, transparent, and predictable legal framework. Such developments will not only safeguard the legitimate rights and interests of copyright and related rights holders but also support the sustainable and responsible development of AI technologies within Vietnam’s music industry.
ASL Law is a leading full-service and independent Vietnamese law firm made up of experienced and talented lawyers. ASL Law is ranked as the top tier Law Firm in Vietnam by Legal500, Asia Law, WTR, and Asia Business Law Journal. Based in both Hanoi and Ho Chi Minh City in Vietnam, the firm’s main purpose is to provide the most practical, efficient and lawful advice to its domestic and international clients. If we can be of assistance, please email to [email protected].
ASL LAW is the top-tier Vietnam law firm for Intellectual Property Services. If you need any advice, please contact us for further information or collaboration.
