GRAI Secures $9M to Make AI Music Social, Not Generative

Music tech startup GRAI has secured $9 million in a seed funding round, aiming to shift the industry’s focus from generative AI tools toward interactive, social experiences. The company, founded by a team of Belarusian entrepreneurs who previously sold their video app Vochi to Pinterest, is positioning itself against the current trend of AI-driven song creation.

The funding round was co-led by Khosla Ventures and Inovo VC. Additional participation came from a diverse group of investors, including Tensor Ventures, Tiny.VC, Flyer One Ventures, the a16z Scout Fund, and several angel investors such as former Pinterest engineer Andrew Zhai, Unreal Labs founder Greg Tkachenko, Rhapsody founder Rob Reid, and MirAI’s Dima Shvets.

Redefining Music Interaction

According to CEO Ilya Liasun, the core problem in the current music landscape is that listening remains largely passive, while social engagement is fragmented. “Discovery is broken, listening is passive, and social context is almost non-existent,” Liasun noted. GRAI aims to address this by focusing on how consumers—specifically Gen Z and Gen Alpha—interact with existing tracks rather than generating new, synthetic ones.

The company has already launched pilot applications, including Music with friends for iOS and an AI-powered playground for Android. These tools allow users to remix songs or alter musical styles for recreational purposes.

Prioritizing Artist Consent

A central tenet of the company’s business model is the “opt-in” approach for rights holders. Rather than launching technology that bypasses labels, Liasun emphasizes that GRAI is in active discussions with industry players to ensure that modified tracks can potentially generate new royalty streams for the original artists.

“We don’t want to share new GenAI slop with the streaming services,” Liasun said. “We actually focus on the interaction part.” The company’s infrastructure includes a specialized “derivatives pipeline” and real-time audio systems designed to preserve the identity of original tracks, ensuring that the technology serves as a creative layer rather than a replacement for human artistry.