Using AI in music composition: which commands create emotion in a work?

THE HANOI TIMES — A powerful new trend has emerged in Vietnam’s digital music space. Songs created and performed by artificial intelligence have moved from experimentation to mass circulation, sparking debate over whether AI should remain a tool or evolve into a genre of its own.

That shift became clear with Say mot doi vi em (Infatuated with you for a lifetime), composed by Ken Quach and Huong My Bong and sung entirely by an AI-generated voice.

The song spread widely online without major promotion. Its success showed how quickly AI music can capture attention and dominate digital platforms.

The sudden rise of “AI music”

What defines AI music in Vietnam is its speed. Amateur creators, young producers and online audiences have embraced it with little resistance. Channels dedicated solely to AI music have expanded rapidly.

TRO Music, for instance, has gained more than 150,000 subscribers, released around 190 AI music videos and attracted over 60 million views. Other platforms, including Ken Quach Official and Nhac Viet AI, continue to grow.

Established artists have also tested AI in their projects. Singer Dan Truong used AI-generated imagery in his 2023 music video Em oi vi dau, though he later admitted the visuals felt “stiff and lacked emotion”.

In October, musician Thai Nguyen released Song of Peace, introducing a virtual singer named AI/T (Ai Trinh), developed after months of research. Singer Ha Anh Tuan and his team applied AI-generated looping animations in the visualized music video Rose to promote the live concert Sketch a Rose.

More recently, Vietnam Green Journey by singer Nguyen Vu, along with In the Next Life, Still a Vietnamese, directed by Tran Thanh Trung, stirred controversy because of their AI-generated visuals.

These examples show that AI has moved beyond experimentation and into the core of music production and presentation.

AI music viral, emotionaless

Supporters argue that AI expands creative access. It helps users complete unfinished ideas, turn emotions into lyrics and explore music without professional training. In that sense, AI lowers barriers and encourages participation.

Many artists, however, draw a clear boundary. Singer-songwriter Hoang Hong Ngoc said AI cannot replace authentic creation. “AI follows rigid commands, so the music lacks heartfelt emotion and the natural flow of memory. The lyrics often sound awkward and forced,” she said.

Singer Tung Duong expressed similar concerns. “AI can now compose, arrange and sing. Many viral songs today come from machines, not people. I enjoy innovation, but I refuse to sing songs composed by AI,” he said.

Composer Nguyen Van Chung views AI as a reference tool rather than a creative force. “At this stage, AI mainly synthesizes data, creates demos and offers a basic listening experience. It still cannot meet the real demands of music creation,” he said.

For Chung, AI produces technology, not art. Art comes from emotion, imagination and lived experience, all elements no algorithm can truly replicate.

Copyright concerns and legal gaps

The rapid spread of AI music has intensified concerns over copyright and artistic value. AI systems can analyze existing works and generate melodies or lyrics that closely resemble those of real artists, increasing the risk of infringement.

At the same time, AI can help protect copyright. Detection tools can identify similarities between songs with high precision, supporting enforcement rather than weakening it.

Technology alone, however, cannot solve ethical problems. As Hoang Hong Ngoc noted, artists inevitably influence one another. Respect for copyright ultimately depends on professional integrity.

From a legal standpoint, lawyer Le Quang Vinh of Bross & Partners said Vietnam’s copyright framework still contains major gaps, especially amid rapid digital transformation and deeper global integration.

He suggested clearer regulations on AI-created works. Songs generated entirely by AI should not receive copyright protection, while works shaped by substantial human creativity may qualify.

Vinh also proposed studying a separate system of related rights for AI products. Such measures could close legal gaps while preserving the core principle of copyright: protection for human creativity.

AI for fun is acceptable, AI for profit crosses a line

Using AI to make music for enjoyment, experimentation and learning poses little ethical risk. Artists have always relied on tools and AI can serve as another creative assistant. The problem begins when AI music enters the commercial market.

Monetizing AI-generated songs relies heavily on uncredited human labor. Algorithms train on vast libraries of existing music created by real composers and performers, often without consent or compensation.

When creators profit from AI music through views, ads, or sponsorships, they benefit from a system built on borrowed creativity. That practice undermines fairness and erodes respect for original work.

Commercial AI music also avoids accountability. Without a human creator, responsibility disappears.

Listeners cannot trace intention, experience, or emotion behind the sound. Commercial art demands responsibility to audiences and to the creative community.

Treating AI music as a product risks reducing music to disposable digital content. AI should remain a supportive tool, not a substitute for earning a living through artistic labor. Once profit becomes the goal, ethical boundaries matter.

AI may dominate platforms and timelines, but emotion remains the foundation of music. As technology advances, Vietnam’s music industry faces a critical choice about whether it can harness innovation without losing the human voice that gives art its meaning.