Music Law and Artificial Intelligence: From Cloned Artists to AI-Generated Works

The rise of artificial intelligence (AI) in the music industry is sparking a revolution, profoundly changing how music is created. This development raises complex legal questions concerning AI and copyright, including related rights. How can we protect the creative rights of artists and composers while simultaneously allowing room for technological innovation? In this comprehensive yet accessible legal overview, we explore key issues regarding AI and music. These include whether AI can legally train on copyrighted materials without consent, TDM exceptions, how various rights organizations (such as Buma/Stemra and Sena) approach AI, the status of AI-generated musical works, the threshold of human creativity required, protection against AI voice cloning via privacy laws and moral rights, contractual implications, new obligations under the EU AI Act, differences between European and American law, and ongoing lawsuits. This article is tailored for artists, composers, music publishers, labels, voice actors, producers, and AI companies seeking clarity on their legal standing.

AI Training on Protected Music and Video Materials: Legal Framework and Debate

Can an AI model in the Netherlands and the EU train on copyrighted material (such as music or video) without permission from the rights holders? Generally, using protected material beyond private use or citation requires permission. Scraping or using data for AI training without permission is typically considered infringement unless a specific legal exception applies.

Buma/Stemra’s Opt-Out Policy

In the Netherlands, Buma/Stemra explicitly uses its opt-out rights, requiring prior consent for TDM on its repertoire, thus ensuring fair compensation for composers and lyricists.

EU AI Act: Transparency Obligations and System Monitoring

The EU AI Act, effective from August 2025, introduces important transparency requirements, obliging generative AI model developers to:

Disclose training data used, including copyrighted music or texts.
Maintain policies ensuring compliance with EU copyright law.
Respect explicit opt-out signals from rights holders during training.

The Act doesn't prohibit using protected material for training outright but enforces transparency and compliance through oversight and penalties.

Composition, Lyrics, and Master Recordings: Different Rights Regimes

Music rights in the Netherlands broadly split into:

A. Copyright: Protects compositions and lyrics, managed by organizations like Buma/Stemra.

B. Neighboring Rights: Protect recordings and performances, managed by Sena.

AI-Generated Compositions and Lyrics: Completely AI-generated works often fail to meet traditional copyright criteria, as human creativity is essential.

Neighboring Rights: It remains uncertain whether AI-generated performances and recordings attract neighboring rights, as these typically rely on human involvement.

Copyright Status of AI-Generated Music

In the U.S., fully AI-generated works explicitly do not receive copyright protection. While Europe hasn't clarified explicitly, the prevailing legal view aligns with this stance—AI-generated works likely fall into the public domain unless there's significant human creativity involved.

Hybrid Creations: Music combining human and AI input may qualify for copyright protection depending on the human creative contribution's significance.

AI Voice Cloning: Personality Rights and Privacy

AI voice cloning technology poses challenges regarding personal rights and privacy. Artists may invoke:

Privacy rights under EU law (Article 8 ECHR).
Personality rights.
Potential trademark and image rights analogously.

The EU AI Act mandates transparency in AI-generated content, aiming to mitigate unauthorized use and deepfake concerns.

Music Contracts in the AI Era

Existing music contracts require updates addressing AI-specific matters, including (1) Explicit licensing terms for AI training; (2) Ownership clarity of AI-generated content; and (3) Liability assignment for copyright infringements involving AI.

Conclusion: Balancing Innovation and Rights—Be Prepared

The intersection of AI and music law presents both opportunities and challenges. Stakeholders should proactively:

Clearly define rights in AI-generated music contractually and update existing music contracts.
Specify permissions (licenses) and restrictions (opt-out) regarding AI training explicitly.
Seek specialized music & AI legal advice to navigate evolving regulations.

By strategically addressing these issues, artists, companies, and AI developers can legally and effectively harness AI innovations, maintaining both creative and commercial control.

Meer lezen

Stanford Light Opera Company: Johann Strauss’ Die Fledermaus

In February 2023, Stanford law scholar Mauritz Kop stepped off the page and into the orchestra pit, joining the Stanford Light Opera Company (SLOCo) for its winter production of Johann Strauss II's Die Fledermaus. He played clarinet—A and B-flat—substage with the company's orchestra across four performances at Dinkelspiel Auditorium. For a researcher whose work increasingly concerns the rights of performers and composers in the age of artificial intelligence, the view from below the stage was more than a hobby; it was a vantage point.

A student company's feminist Strauss

SLOCo's staging, directed by Nicolle Hendzel '23 in an English translation by Marcie Stapp, reimagined a ballroom farce as a deliberately feminist work, placing the characters who orchestrate the plot's great prank—and who hold power within it—in the hands of its women. The company is a student-run organization with roots in the Stanford Savoyards, a Gilbert and Sullivan society founded in 1973 and rebranded as SLOCo in 2016. The operetta itself dates to 1874, when it premiered at the Theater an der Wien in Vienna; nearly a century and a half later, a student orchestra gave it a contemporary edge.

Where music meets the law of AI

Kop is a Stanford Law School fellow and a practicing classical musician on piano and clarinet, and the two roles inform one another. His scholarship on copyright, authorship, and the public domain carries the weight of someone who has sat among the performers the law is meant to protect. That sensibility runs through his analysis of cloned voices and machine-generated compositions, where familiar copyright questions—who is the author, who is paid, and what enters the public domain—have been made newly pressing by generative systems. A scholar who has taught music law at the Royal Conservatoire in The Hague has good reason to keep one foot in live performance.

Why the arts belong in technology-law writing

Technology law has a tendency to treat the arts as something to be regulated rather than understood. An evening in a student orchestra reverses that order. The discipline of a rehearsed part delivered in real time, the collective labor of an ensemble, and the precarious economics of a production staged for love rather than profit are precisely what the law of music exists to serve. As AI begins to imitate the performer's craft at scale, the argument for a creator-aware approach to authorship and remuneration only sharpens—an argument that runs throughout Kop's scholarship on responsible technology across law, music, and policy. The clarinet, on this telling, is not a distraction from the research—it is a reminder of whom the research is for.