Technology news and discussion on TalkFuse.
Key Principles for Ethical AI Use in Music Creation
A lawyer specialising in entertainment law has set out principles for ethical use of artificial intelligence in music creation, arguing that licensing deals between rightsholders and AI firms must be built on artists' consent, fair pay and transparency. Writing in an opinion piece, John Meller of Manatt, Phelps & Phillips says the industry is at a turning point as companies including Suno, Udio, Klay and Spotify pursue licensing arrangements to legitimise their AI offerings, and warns the sector risks repeating the mistakes of the piracy era if it fails to bring creators along with it.
Meller argues that songwriter and artist approval for AI training and licensing should be treated as non-negotiable, regardless of how existing recording and publishing contracts are worded. He cites Universal Music Group's reported deal with Spotify, under which UMG artists and writers must actively opt in, as a positive example, while criticising other labels for striking AI partnerships without seeking creator permission or for inserting clauses claiming perpetual rights to AI-generated voice and likeness. He notes that legal action against AI firms that train without permission is already under way, including a lawsuit his own firm filed against Anthropic on behalf of BMG.
In short
- Lawyer proposes ethical principles for AI music licensing deals
- Calls for mandatory artist/songwriter consent and fair compensation
- Cites UMG-Spotify opt-in deal as a positive model
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Artists created the works that train these systems and deserve meaningful control over their use and fair compensation for it. Transparency and opt-in arrangements, as exemplified by Spotify's UMG deal, demonstrate that legitimate licensing is genuinely possible. The industry must learn from the piracy era—creators must be brought along as partners, not excluded from decisions about their work. Permission-based systems with fair payment structures protect both the creative workforce's interests and the industry's long-term sustainability.
The case against
Broad opt-in requirements could significantly impede AI development and the innovation benefits these tools offer creators of all sizes. Existing recording and publishing contracts may legitimately cover AI training uses, and companies negotiated those agreements in good faith based on established rights. Training on published works may constitute fair use for transformative purposes, a legal principle foundational to copyright law. Moreover, AI music tools are democratising creation itself, enabling emerging and smaller artists to produce professional-quality work they couldn't otherwise afford, and overly restrictive consent mechanisms risk pricing out innovation that benefits the broader creator economy.
Explained simply
Artificial intelligence systems are now being developed to generate music, often by learning patterns from large collections of existing recordings and compositions. The question of how these systems should be trained – specifically, whether artists whose work is used to train them should be asked permission or compensated – has become a significant dispute across the music industry. The technology is advancing quickly whilst the industry is still negotiating how to approach these issues.
The music industry is currently negotiating licensing agreements between AI companies and record labels, music publishers, and rightsholders to establish frameworks for the technology. Different companies and labels have taken different approaches, with some requiring artists to opt in before their work is used in AI training, whilst others argue that existing recording or publishing contracts give them the right to do so without additional artist approval. This reflects earlier industry disputes around digital distribution and music copying technologies.
At stake is whether musicians and songwriters will have control over how their work is used to develop AI systems, and whether they will receive compensation if that happens. The approach now established for licensing and consent could set industry standards that shape how broadly AI training can use creators' work, and what rights and payments creators retain. Legal disputes are already under way, with some music organisations challenging certain AI firms' practices.
AI Art Cricket Culture Entertainment Music Sport Technology