AI music tools have exploded in popularity, with startups promising instant songs made from scratch using artificial intelligence. But a new music lab called GRAI is taking a different bet: most people don’t actually want to “become” music producers. They want to do lighter, more social things with music they already love, like remixing a track, swapping it with friends, or changing its vibe just to see how it sounds.
That shift matters, because it also raises a key question: who gets to decide whether a song can be remixed or transformed in the first place? GRAI’s answer is clear—artists and rights holders should be in control.
GRAI has now secured $9 million in seed funding to build AI-powered music experiences designed around participation, not pure generation. The company was created by Belarusian founders who previously built the video creation app VOCHI, later acquired by Pinterest. Now, they’re applying that product intuition to music, experimenting with apps that let everyday listeners interact with songs in more playful ways.
So far, GRAI’s early products include Music with Friends for iOS, built around remixing and sharing, plus an AI music playground for Android. The company says these early releases are part of a broader learning strategy: ship quickly, see how people actually want to engage, and use feedback to shape what comes next.
GRAI co-founder and CEO Ilya Liasun, based in Poland along with much of the team, believes music is one of the last huge consumer categories that hasn’t fully embraced a “creator-first” evolution. In his view, key parts of the music experience still fall short for modern audiences: discovery can feel broken, listening is often passive, and the social layer around music is surprisingly thin.
Rather than framing AI as a threat that replaces artists and labels, GRAI argues that AI can unlock new forms of engagement that don’t flood the internet with low-quality auto-generated songs. The focus is on interaction—tools that let fans participate without pretending they’re professional creators.
That approach is aimed especially at Gen Z and Gen Alpha, where music discovery is driven less by traditional browsing and more by culture: friends, fandoms, and short-form video. Many of these listeners don’t want to sit down and craft music from scratch. They want to join in, react, remix, and share moments.
Behind the scenes, GRAI has built its own infrastructure to make this kind of social music interaction possible. The company developed a “taste and participation graph” to understand how people connect with music and with each other. It’s also working on what it calls a derivatives pipeline and real-time audio systems designed to preserve the identity of the original track while still allowing transformations like style changes or remixes.
Importantly, GRAI says the goal is to make this activity legal and rights-respecting by working with artists and labels, not bypassing them. Liasun has emphasized that the company isn’t trying to push “genAI slop” into streaming services. Instead, it wants to keep the action inside interactive experiences where fan creativity can happen responsibly.
A key part of GRAI’s model is opt-in and opt-out control for artists and rights holders. The company says it’s approaching labels first before building too far ahead, with the intention of creating a future system where owners give permission before their music is integrated. Liasun hasn’t shared whether deals are already in place or with which partners, but he describes consent as a core principle.
If social remixing and track transformations take off, GRAI believes it could also become a new discovery engine—helping listeners find artists and songs beyond the usual short-form video feeds and major platforms. And because remixed or modified versions could potentially create new royalty streams, the company sees a path where fans get more ways to participate while artists and labels get paid.
For now, GRAI is intentionally treating its first apps as a testing ground. The team is looking for honest consumer reactions, including negative feedback, to quickly learn what resonates and what doesn’t.
GRAI was co-founded by CTO Dima Kamarouski and President Andrei Avsievich. The $9 million seed round was co-led by Khosla Ventures and Inovo vc, with participation from Tensor Ventures, Tiny.VC, Flyer One Ventures, a16z Scout Fund, and several angel investors including Andrew Zhai, Greg Tkachenko, Rob Reid, and Dima Shvets.
