Independent music publishing organisations IMPEL and IMPF have issued a licensing framework for generative AI.
According to the announcement, the licensing framework is based on key principles on AI that will ensure song copyrights are properly valued.
The framework is based on a number of key licensing considerations including revenue allocations between master and publishing rights; payments for past use, training outputs and future exploitation; and mechanisms to establish trust and transparency.
IMPF and IMPEL are now calling on CMOs and other licensors to support the principles to create a fair and sustainable future for the music industry.
With the music business at a new inflection point, it is crucial that we take a strong stance that will set the right precedent for the industry going forward
IMPEL CEO Sarah Williams said: “The principles we have established here have been designed to protect songs and songwriters, and ensure they are properly valued. With the music business at a new inflection point, it is crucial that we take a strong stance that will set the right precedent for the industry going forward.”
IMPF president Annette Barrett said: “IMPF exists to give independent publishers and the songwriters we represent a collective voice, a mission that has never mattered more, as our industry navigates disruptive change and new rules are being written. The song needs protecting as the keystone of it all.”
The full framework can be found below.
Joint Statement: IMPEL and IMPF on fair licensing for generative AI
IMPEL and IMPF believe that generative AI licensing must be founded from the outset on fair, transparent and sustainable principles. The independent music publishing community is ready to engage constructively with AI companies, technology providers and licensing partners, but emerging AI business models must not repeat or deepen the structural imbalances that have shaped music publishing in the digital market. This is especially important in a new context in which the song plays an even more significant role.
One principle is non-negotiable: the song must be properly valued and, absent other salient factors, at least equally with the recording.
In generative AI, songs, lyrics, melodies and compositions are central to the training, prompting, generation, output and commercialisation of AI music products. Whilst AI generation can take place in various forms that must be treated on their merits, any licensing model that, ab initio, treats publishing as a minor share of value while allocating the overwhelming majority of revenues to master rights is unacceptable.
After years of addressing unequal revenue splits in streaming, the independent publishing community sees generative AI as an opportunity to establish a fairer framework, not another market in which outdated assumptions about the value of masters and compositions are carried forward.
Transparency must be embedded throughout the licensing process.
Generative AI licensing must reflect the distinct and essential contribution of songwriters and publishers, and must be clear in scope. Limited pilots, closed environments or specific product models must not be used to create broad precedents for future exploitation. Rightsholders must understand what is being licensed, how works are being used, how attribution will operate, what deductions are being applied, and how revenues will be calculated and distributed.
Technology can play an important role in building trust, supporting clear licensing terms, robust reporting obligations and rightsholder oversight.
For this reason, IMPEL and IMPF are continuing to assess the capabilities of attribution technologies and are currently conducting three pilots with different companies. These pilots are examining how technologies can support the identification, tracking and attribution of musical works across AI training, generation and outputs, as well as the reporting necessary for accurate remuneration.
We support innovation. We support responsible experimentation. We support constructive licensing discussions. But innovation cannot be built on unclear terms, unfair economics or licensing models that undervalue the song.
As representatives of independent music publishers, IMPEL and IMPF call for generative AI licensing frameworks that uphold the following principles:
– The song must be properly valued in all generative AI licensing models.
– Revenue allocations between master and publishing rights must be fair and balanced but in the absence of any other salient factors, generative AI training and exploitation should value the song at least equally with the recording.
– Payments for past use, training, outputs and future exploitation must be clearly distinguished and fairly treated.
– Any deductions, costs or revenue calculations must be transparent, justified and applied fairly.
– Licence scope must be specific and must not create unintended pathways for broader future uses.
– Mechanisms to establish trust and transparency in the attribution and valuation of original works used in AI training and output must be agreed with rightsholders.
– Downstream uses of AI-generated music must yield appropriate royalties for the songs on which AI models are trained.
Generative AI is shaping the next phase of the music economy. The rules being set now will have long-term consequences. IMPEL and IMPF urge all parties to work with independent publishers and songwriters to build a licensing framework that recognises the true value of the song and creates a fair, sustainable market for the future.
