Orfium awarded funding to lead research into copyright detection in AI music

Orfium, the global technology company of music rights management, has been awarded a major grant from the European Commission to lead a research project that will tackle how to detect and attribute the use of copyrighted works in AI-generated music.

The funding awarded to the Company is part of a €7.5 million project, named AIXPERT, by the EU’s Horizon Europe programme under the Explainable and Robust AI initiative, which aims to improve transparency and accountability in artificial intelligence systems across sectors. Orfium will work as part of a pan-European consortium and was one of only three proposals selected from 135 applications to obtain the grant.

The consortium, which includes Sorbonne University, Athens Research Centre, Amsterdam University Medical Centre, Barcelona Supercomputing Centre, the University of Barcelona, Novelcore, Furhat Robotics, Kyklos Ltd., Workable, Infinitivity Design Labs, ITML, Martel Innovate, Philips Consumer Lifestyle, the University of Groningen, the Vector Institute for Artificial Intelligence and the French National Centre for Scientific Research, has been tasked with developing AI technologies that can explain their decision-making processes. It will seek to deliver a comprehensive, human-centric AI framework grounded in FATE principles: Fairness, Accountability, Transparency, and Ethics. The project will create AI systems that are transparent, traceable, and inclusive by design.

Pioneering explainable AI in music

Orfium is the only member of the consortium leading the application of this technology in the music industry, using explainable models to identify when and how AI-generated music incorporates elements of existing human-made compositions. Detecting these instances is the first step in enabling accurate attribution and compensation for original songwriters, composers and rightsholders.

The technology pioneered by this project will help to ensure that those who create the original human compositions are not left behind and can benefit financially, at a time when the Gen AI music and audiovisual content market is projected to grow from €3 billion now to €64 billion in 2028, according to CISAC (International Confederation of Societies of Authors and Composers).

The project is believed to be the first research of its kind at this scale, and it could pave the way for AI-era copyright regulation, licensing and monetisation standards that protect songwriters, composers and publishers in the future.

Rob Wells, CEO, Orfium said: “Coming at a crucial time in the development process of AI, this project is hugely significant for the music industry and fits closely with our mission to apply AI for the good of the sector. As generative AI reshapes the entertainment landscape, we are building the infrastructure to ensure creators remain at the centre of that evolution. To be selected by the European Commission is a strong validation of Orfium’s position at the forefront of innovation in music rights management. It aligns with our commitment to leverage AI to tackle the entertainment industry's most complex challenges.”

The three-year project officially began on 1 June 2025. The research funding is part of the EU’s drive to remain competitive in the AI sector in the face of increasing global competition.

Haris Papageorgiou, AIXPERT Coordinator said: “We are at a critical inflection point where AI's tremendous potential can only be fully realised if we get the fundamentals right from the ground up. This isn't just about building more powerful systems – it's about building AI that people can actually trust and understand. When an AI system makes a recommendation about someone's medical treatment or hiring decision, that person deserves to know not just what the decision was, but why it was made and whether it was fair. AIXPERT with its framework embeds these FATE principles directly into the architecture itself, so transparency and accountability are not afterthoughts – they are the foundation. We are essentially creating AI that can explain itself in human terms while maintaining the rigor and performance that makes it valuable in the first place.”