Photo Credit: Jordan Whitfield
The Independent Music Publishers International Forum (IMPF) has proposed four ethical guidelines for AI developers.
What are the rules for AI? Digital Music News is assembling a critical forum of music industry leaders to address this very topic. Join major labels, publishers, artists, and leading music tech players on October 25th in Los Angeles — tickets here.
Composed of 200 of the world’s most significant independent music publishing companies, IMPF is the global trade and advocacy body for independent music publishers worldwide. Members include Concord Music Publishing, Big Machine Music, Downtown Music Publishing, and Reservoir.
The IMPF informed DMN via email today that it has released four ethical principles to foster higher transparency among songwriters, music publishers, and developers that are training AI music models.
IMPF’s latest move to ‘guide’ generative music AI comes in the midst of the industry’s rapid acceptance of AI, alongside frequent copyright infringement litigation and discourse surrounding AI models’ lack of transparency. Despite the evident boom of generative AI acceptance and practical use within the music industry, there have been calls to ensure AI developers maintain accurate records of music utilized in training AI. Organizations in the EU had also recently called for mandatory artificial intelligence training disclosures, saying, “AI innovation and effective copyright protection are not mutually exclusive.”
Against this backdrop, the IMPF proposed four principles that apply specifically to generative AI. The move aims to ensure the ethical use of music during the training of AI applications by developers, and create an environment of collaboration among all parties involved.
In a document detailing the guidelines, IMPF states, “The protection of human writers’ copyright and livelihood should be explicitly acknowledged and provided for in any AI-related activities, commercial negotiations or legislative initiatives. Applying generative AI should be based on principles of legality, accountability, and transparency.”
- Seeking express permission for the use of music in the machine training process: “Superseding social interest exists which would justify an exception to this human right of intellectual property which protects not only the writers’ economic interest but also the expression of their creative personality.”
- Keeping records of the musical works used in the machine training process: “The process of ingesting existing musical works constitutes the best opportunity to know the creative elements used by AI applications. It is relevant for any licensing, should the writer or rightsholder choose to allow the use, to ensure that any remuneration can be accurately distributed to the writers or rights holders.”
- Labeling human created and AI generated works: “We suggest a clear differentiation between human-created musical works which benefit from copyright protection and machine-generated music which does not. Such clear demarcation of AI-generated musical works will ensure a level playing field for human-created music while protecting consumer choice.”
- Delineation between assistive AI and fully generative AI application: “We strongly urge differentiating between human creation and technical generation, in particular by clearly labeling AI-generated musical works as such.”
Annette Barrett, President of IMPF, said, “We have established these four key ethical principles to strike a careful balance between progress and protection, to assert creative rights, and, ultimately, to forge a healthier relationship between the creative and technology industries.”
Barrett also emphasized that despite the debate — and often panic — about the rapid development of artificial intelligence and its implications for the creative industries, the technology is here to stay.
Barrett added, “[AI] will only get more sophisticated. In many cases, it will actually enhance our work and lives. We should not fight these advancements, but it would be negligent to give tech developers free reign when it comes to the use of artistic human work — which carries its own irrefutable, intrinsic value — to enable machine learning.”
IMPF’s new principles are one of the industry’s more extensive attempts to control artificial intelligence. The guidelines surface on the first day of the second IMPF Global Music Entrepreneur and Creative Industry Summit in Palma de Mallorca, Spain. The event is taking place on October 8th to 11th, and is expected to have over 340 music industry professionals in attendance today.
