As artificial intelligence (AI) companies, creatives, and copyright holders clash, Canada’s Musical AI has teamed up with another early-stage music technology startup to build what it says will be a “completely legal and licensed” AI song generator.
The company has partnered with India-based Beatoven.ai to develop the new service, which will combine Musical AI’s rights management platform with Beatoven.ai’s AI song generation software. They say the AI song generation model, which they plan to launch in the second half of 2025, will be trained on over three million songs, loops, samples, and sounds—all with permission from and compensation for rights holders.
Musical AI aims to help rights holders like musicians and record labels get paid for licensing their music to generative AI companies.
AI has come for the music industry: this partnership comes amid a rise in AI-generated music. Many musicians have pushed back against more predatory applications of AI to music, while others have embraced the technology as a tool for supporting their work. Meanwhile, some of the companies powering the creation of AI songs have faced lawsuits from record labels alleging that they have scraped copyrighted songs to train their models without consent.
“Generative AI is already playing a significant role in music creation,” Musical AI co-founder and CEO Sean Power told BetaKit over email. “Rather than fighting it, we believe that it’s important to lead by example in creating tools that honour artists’ contributions and compensate them fairly.”
Beatoven.ai is among the growing group of AI companies that help users generate full songs based on text, audio, and video prompts. It has gathered traction as a tool for making royalty-free background music for videos, podcasts, audiobooks, games, and other uses: Beatoven.ai claims it has helped over 1.5 million users create more than six million tracks to date. It claims to provide musicians with “equitable compensation” for contributing their music.
Enter Musical AI. Founded by Power and two other musicians with experience in tech—COO Matt Adell and CTO Nicolas Gonzalez—the startup has built rights management, licensing, and attribution tech for AI music. “Our aim was to develop a system where rights holders could securely license their music for AI training with proper attribution and compensation, and where AI companies could access high-quality data ethically and legally,” Power said.
“We launched Musical AI in late 2023 after recognizing the tension between advancements in AI-generated music and the rights of musicians and record labels,” Power said. “While lawsuits against AI companies for using copyrighted music without consent were a factor (and it was obvious they were coming), our decision was largely driven by a deeper understanding of the need for a collaborative solution.”
RELATED: Music collaboration startup BeatConnect raises $2.25 million CAD
Musical AI’s software helps intellectual property (IP) holders like musicians and record labels get paid for licensing their musical data to generative AI companies. The startup aims to offer music rights holders access to new revenue streams and help them retain control over how their IP is used, while also protecting AI companies and their clients from lawsuits. Musical AI already has agreements with rights holders like Symphonic and Kanjian.
To date, Musical AI has secured $1.3 million USD in equity funding as part of a pre-seed round led by Halifax’s Builder Ventures that it plans to fully close by early 2025. The round saw support from a group of angels including British musician Tommy Danvers of Ministry of Sound and Right Said Fred. The startup intends to raise seed financing during the first quarter of 2025.
“We believe with this partnership, we will set the way forward for how business models need to be built in AI with the rightsholders being compensated for the data the models are trained on,” Beatoven.ai founder and CEO Mansoor Rahimat Khan said in a statement. “We have historically been adopting this model in direct partnerships with independent artists and by joining hands with Musical AI we will build a sustainable revenue-sharing model using their attribution technology.”
RELATED: Beatdapp unveils $22.9 million CAD, partnerships with Universal Music Group, SoundExchange, Napster
Though this partnership with Beatoven.ai, Musical AI will provide data licensing, attribution of generated outputs, and payments to rights holders, who the company said will receive “an appropriate share” of the model’s revenue based on usage, likening it to how they are paid when music is streamed on a commercial service. This AI song generation model will be trained on the existing rights holder catalogues of Musical AI.
“As musicians ourselves, we view AI as a valuable addition to the industry—one that opens up new possibilities without diminishing the role of the human at the centre of the creative process,” Power said. “It’s crucial to us that AI models respect and acknowledge the contributions of human creators, which is why proper rights management and attribution are at the core of what we do.”
Power said it is up to artists and listeners to shape the future of music and AI. “Whether AI is used as tools to help song creation, or to generate background music for content, or even to create entire songs without human input—consumer preferences will drive these developments,” he said.
Musical AI hopes to ensure that it is “used in a way that benefits everyone.”
Feature image courtesy Unsplash. Photo by Marcela Laskoski.