When Machines Compose: AI and the Future of Music
Not long ago, the idea of a machine generating convincing music across multiple styles and genres in real time was unimaginable. Today, it is a commercial reality — and as a composer, musician and UNESCO cultural expert, I find it impossible to look away.
The rise of AI-generated content has touched almost every creative discipline, and music is no exception. What began as an academic curiosity has become a rapidly expanding industry, with dozens of platforms now offering tools that can produce lyrics, melodies, harmonies, and full arrangements — often for free, and often with startling sophistication. The development of AI-generated music cannot be separated from the broader explosion of AI-generated content across all fields.
What is the impact of artificial intelligence on the future of music and singing? What challenges does human creativity face in the era of machine-generated content? These are not rhetorical questions, they demand urgent answers from artists, institutions, and governments alike.
I write this not only as a composer or musician, but also as a cultural expert who has spent several years to help shape the international response working directly with UNESCO on the intersection of digital environment including AI and cultural policies — first as a member of the Reflexion Group on diversity of cultural expressions in the digital environment (2023–2024), then as chair of the Independent Expert Group on AI and Culture (2025), whose 80-pages report I presented in Barcelona at MONDIACULT 2025.
What follows draws on that dual vantage point: the practitioner who composes and performs, and the expert who has sat at the table where the international policy response to this transformation is being forged. And the question is not simply whether AI can make music, it demonstrably can. The deeper question is what this means for composers, performers, technicians, listeners, and the very concept of musical authorship?
Let's go see all of this.
A Historical Turning Point
AI represents a historic inflection point in human creativity. Music is entering a post-AI era — a shift comparable to the introduction of musical notation in the 10th century, Gutenberg's printing press in the 15th century (which democratized the distribution of musical scores), or the arrival of electronic sound in the mid-20th century. The scale of disruption is real, but so is the historical precedent.
Yet there is a crucial difference between this shift and those that came before. Each previous revolution transformed one specific dimension of musical life: notation changed how music was written and preserved; the printing press changed how it circulated; electronic sound changed how it was performed and recorded. The AI revolution, by contrast, is penetrating all of these dimensions at once, and at a speed that leaves little time for gradual adaptation. What took centuries to absorb, music communities are now asked to process within years, even months. This compression of historical time is itself one of the defining challenges of the current moment.
A Revolution Already Underway
Until few years, it was unimaginable that a machine could produce music across multiple styles simultaneously, with genuine conviction. Today, some AI platforms can generate content that fools even the trained human ear, creating emotions and sensations comparable to those achieved by human artists. The development of AI-generated music cannot be separated from the broader explosion of AI-generated content across all fields.
Among the leading platforms shaping this landscape:
Genuine Opportunities
The rise of AI music tools brings meaningful positive possibilities, and the most fundamental of these is Democratization of music production. For the first time in history, anyone — regardless of musical training, technical knowledge, or financial resources — can generate lyrics, melodies, harmonies, and fully produced tracks simply by describing what they want. A teenager in Algiers, a filmmaker in Rio, a poet in Amsterdam: all can now produce music that would previously have required years of study, expensive studio time, and an entire network of professional collaborators. The barrier between having a musical idea and realizing it has, in practical terms, collapsed.
This shift is not without precedent. Some twenty years ago, the art world had already witnessed a comparable disruption — in photography. The arrival of digital cameras and then smartphones did not merely improve photography; it abolished the entire economic and professional ecosystem that had grown around it. Street photographers, portrait studios, film manufacturers, darkroom technicians, camera repairers, photo development shops: entire trades that had existed for over a century disappeared within a decade, not because they became unnecessary in any abstract sense, but because the technology made their function accessible to everyone for almost nothing. Today, billions of people are photographers — and yet the profession of photography, as it was understood in 1990, barely exists. True professionals must either adapt, become more creative and reinvent their art and craft, or quietly leave the scene.
The parallel with music is striking: democratization and disruption are two sides of the same coin, and the speed of one determines the violence of the other.
The Transformation of the Music Industry
Beyond this democratization of music production, AI also opens genuinely new creative territories: artistic forms that were previously impossible, new tools that allow musicians to expand their compositional horizons, and a more interactive and personalized relationship between artists and their audiences. These are not trivial contributions. But they must be weighed honestly against the costs — which, as the history of photography suggests, will fall most heavily on those who made their living from the skills that technology is now making redundant.
The 20th-century music industry operated through a clear sequential Music Economy Value chain: creation, training, recording, production, distribution, and marketing. The internet and social media began compressing this model at the turn of the century, allowing artists to reach audiences directly. Today, AI is penetrating every single stage of this chain — from composition and arrangement to production, mixing, mastering, distribution, and audience targeting through intelligent algorithmic recommendations.
What makes this shift qualitatively different from previous ones is the nature of AI's intervention: it does not insert itself into one stage of the chain and leave the others intact. It operates across all of them simultaneously, all at once and often through a single platform. In the 20th-century model, a composer depended on a studio, which depended on a label, which depended on a distributor: each link was a distinct profession, a distinct community of expertise. Today, a single AI tool can bypass every one of those links in a matter of seconds. The value chain in music industry has not just been broken but it is totally disrupted and disordered.
Real Dangers for Music Professionals
It is precisely this simultaneity and this acceleration that makes the current shift so threatening. Because all stages of the music industry are being disrupted at the same time, and because the pace leaves almost no room for gradual adaptation, the risks for musicians and technicians are not abstract or distant: they are already unfolding at every level of the industry:
It should be noted that this upheaval will, in my opinion, have little effect on the names and major artists already established — those who have built significant careers and forged distinct artistic identities. Their audience loyalty, their brand, and their creative authority are not easily replicated by an algorithm. But for the wider ecosystem of active musicians and technicians — the less well-known, the younger, those who do not yet have solid networks or a recognized name — the impact will be significant. It is precisely this silent majority of the music world that bears the full weight of the disruption.
The Ethical Crisis
One of the most serious issues raised by artists, experts, and copyright protection bodies is the "Unauthorized use of copyrighted works to train AI systems" — without the consent or compensation of the creators whose music, texts, artworks, and research were used. The legal frameworks governing this remain outdated and unresolved, and the financial entitlements of creators in the digital environment are still far from clearly established.
As early as October 2023, France's SACEM (the Société des Auteurs, Compositeurs et Éditeurs de Musique, one of the world's oldest and most influential copyright societies, founded in 1851) exercised its formal opt-out right against AI data-mining. The decision was unambiguous: any entity developing AI tools would henceforth require prior authorization from the SACEM before conducting data-mining activities on the works in its repertoire — a catalogue covering some 96 million protected works. The SACEM's position rested on a clear principle: AI systems are trained on vast databases that inevitably include protected works, and this constitutes an exploitation of creators' rights for which they must receive just compensation.
Another reaction from the artistic community was swift and massive, coming from US. In April 2024, more than 200 prominent artists signed an open letter directly targeting AI music generators, calling their practices predatory and destructive to human creativity. Among the signatories were Billie Eilish, Nicki Minaj, Stevie Wonder, and Katy Perry — joined by the estates of Bob Marley and Frank Sinatra. The letter was not a vague expression of concern: it was a direct demand for the industry to halt the unauthorized exploitation of artists' work as training data without consent, credit, or compensation.
The American Federation of Musicians (AFM) and the Music Workers Alliance (MWA) also joined in support of the lawsuits, the latter stating plainly that AI platforms were "forcing musicians into a training role to which they never consented."
These cases are not merely legal disputes: they are the first serious institutional attempt to define the rules of the road for AI development in music — and their outcomes will shape the entire field for years to come.
What Institutions and Governments Must Do
In a world of accelerating technological change, the international community faces the challenge of balancing technological innovation with the preservation of cultural heritage. "Digital sovereignty" (a nation's capacity to protect its own cultural content in the digital space) is increasingly at stake.
Several frameworks are emerging in response: the EU's General Data Protection Regulation (2016), national laws in the UK, US, Brazil, South Korea, and the EU AI Act (2024) — the first international legislative text specifically aimed at regulating AI use while protecting cultural heritage and cultural actors.
UNESCO has been working on this since 2018 through its 2005 Convention on the Protection and Promotion of the Diversity of Cultural Expressions. In December 2021, the organization adopted a landmark global Recommendation on the Ethics of Artificial Intelligence — the first of its kind worldwide — establishing ethical frameworks across culture, education, environment, communication, data, and gender equality.
In 2023 and 2024, UNESCO established an expert group of 18 international experts (of which I was a member) focused on four priorities: linguistic diversity in digital cultural content, discoverability of national and local content, platform transparency on cultural matters, and the impact of AI on cultural and creative industries.This work culminated in the publication of 11 Recommendations of the reflection group on the diversity of cultural expressions in the digital environment (32 p.), which were presented in Paris, February 11-14, 2025, at the Intergovernmental Committee of the 2005 UNESCO Convention, in front to 161 member states.
In 2025, another group of 12 independent experts was mobilized to study the issue of AI in culture. I had the honor of chairing this group and directing the drafting of the 80-page/8 chapters Report of the Independent Expert Group on 2025 Artificial Intelligence and Culture, which I presented in September 30, 2025 in Barcelona at MONDIACULT España 2025.
AI carries enormous potential that is increasingly reshaping music and song, restructuring the value chain, and opening genuine new opportunities. But the risks to musical professions — in terms of practice, creativity, ethical integrity, and moral rights — cannot be ignored.
Societies, institutions, and governments must not remain passive spectators or mere consumers in the face of the hegemonic model imposed by large technology corporations. They must become active players — ensuring that the use of AI strikes a genuine balance between encouraging technological innovation and protecting human creative practice.
"The challenges we face today are not merely obstacles — they are opportunities to rethink how technology can be integrated positively and sustainably into our world."
Salim Dada, Al-Araby Al-Djadeed, August 30, 2024
A Tool, Not a Replacement
My conclusion, as music practitioner, cultural actor & tech observer, is measured rather than alarmist. AI music generation is a powerful tool — genuinely useful for prototyping, for inspiring harmonic ideas, for making music accessible to people without formal training.
But the tool is not the art. Music that matters — music that moves us across decades and cultures — has always been rooted in human experience and human intention. AI can approximate that experience with increasing fidelity. It cannot originate it.
For composers and musicians, the imperative is not fear but literacy. Understanding these tools — their capabilities, their limits, their ethical implications — is now part of professional responsibility. Remember the photographers who refused to digitize their work, or, long before them, the silent film actors who refused to use their voices in the first talking pictures? They simply perished despite their talents, expertise, and aura at the time.
The question is not whether AI will change music. It already has. The question is whether human artists will shape that change, or simply be shaped by it.
More :