Musical AI, formerly Somms.ai, unveils new platform for ethical use of copyrighted material in generative AI
Musical AI, formerly known as Somms.ai, has launched a new rights management platform designed to allow proper attribution and compensation for the use of copyrighted works in AI training and generation.
The platform addresses a growing concern in the AI space: the lack of a system for attributing and licensing copyrighted material used to train AI models, said Musical AI on Wednesday (June 12). This can lead to situations where AI-generated music or art borrows heavily from existing works without proper credit or payment to the original creators.
Musical AI says its new platform will serve as a “secure go-between” for rightsholders (labels, publishers, artists) and generative AI companies. The Musical AI team comprises experienced entrepreneurs, AI researchers, and music industry veterans, “all of whom care deeply about the fate of data and art in the AI age.”
The company’s advisory board includes Vickie Nauman, a music licensing expert and founder and CEO of LA-based boutique music tech consulting and advisory firm CrossBorderWorks, and Alastair Croll, a technology entrepreneur.
“There are a few futures ahead of us; one is a future where inputs are seen as interchangeable commodities without distinct value. This is the future currently being promoted by some AI companies.”
Sean Power, Musical AI
“We have these incredible advances in AI outputs that require valuable human-created input. There are a few futures ahead of us; one is a future where inputs are seen as interchangeable commodities without distinct value. This is the future currently being promoted by some AI companies,” said Musical AI co-founder and CEO Sean Power.
“In another, arguably better future, we as humans insist that inputs are important, that music, art, ideas, words, and the human labor required to create them have value. If you value this work, then logic dictates that an attribution platform must exist. We are that platform.”
For rightsholders, the platform provides tools to monitor the use of their works, request takedowns of unauthorized content, and set limitations on how their material can be used in AI generation. For AI companies, the platform will allow them to gain access to a repository of licensed data for training their models, while receiving reports on the usage of each source material within their generated outputs.
Musical AI says its platform is able to trace the influence of training data on AI-generated outputs. It can determine the percentage contribution of each source material to a specific AI creation, allowing for fairer attribution and compensation for rights holders when their work is used by AI.
“The current problem in the marketplace is that many AI companies agree that the people who own music and IP should be paid. However, rights holders don’t have mechanisms in place to control and monitor usage.”
Matthew Adell, Musical AI
Matthew Adell, co-founder and COO of Musical AI, said, “The current problem in the marketplace is that many AI companies agree that the people who own music and IP should be paid. However, rights holders don’t have mechanisms in place to control and monitor usage.”
“We’ve created a platform that allows musical assets to exist in a secure environment and that allows us to attribute every single generation. This lets rights holders and AI companies do business fairly and legally.”
The company says it is one of the first to gain certification from generative AI watchdog, Fairly Trained. The non-profit, founded by former Stability AI executive Ed Newton-Rex, provides a “Licensed Model” certification, similar to organic food certifications, that evaluates and certifies AI models based on their training data sources.
Music Business Worldwide
