TuneCore has been at the forefront of the DIY artist movement since 2006 and recently struck a first of its kind partnership with musician Grimes.

So, we sat down with CEO Andreea Gleeson to discuss this new project and the potential opportunities and challenges surrounding AI and music.

You recently announced a partnership with GrimesAI. How did that come about?

What really brought AI to the forefront in music was Ghostwriter putting out that song using Drake and The Weeknd’s voices. That was problematic because one, neither of those artists provided consent for their voices to be used, two, they had no input or control in the creative, and three, they were not participating in the revenue.

Fast forward a few weeks later and Grimes came out and said, “I’ve already been training AI on my voice, here’s the technology and if you want to use it you just need to pay me 50% commission.” I was in Ibiza for IMS at the time and Grimes’ manager and CreateSafe founder Daouda Leonard was also there.

We synced up because I felt that what they were doing was really positive in setting the principles around engaging with AI responsibly, and I knew we could create a way to power what they wanted to do at scale. And thus, the partnership began.

This is the first scale initiative we’re doing with AI and it’s still very much in a pilot phase. We’re working to optimize it further to make it more available at scale, and for more artists to participate in. But really, we wanted to test how we do this responsibly.

This is the first scale initiative we’re doing with AI and it’s still very much in a pilot phase. We’re working to optimize it further to make it more available at scale, and for more artists to participate in. But really, we wanted to test how we do this responsibly.

How does it work?

Any TuneCore artist that wants to use Grimes’ voice is directed to Elf.Tech. She trained the AI on her voice and music that she owns 100% which is key because as we start to think about creating a better structure for monetization, it’s really important to have the rights to train the AI with that music.

From there the artist can transform an audio file into Grimes’ voice or sing directly into the app and then export it. Then once the music is ready, they would go through the usual process in TuneCore, listing GrimesAI as the primary or featured artist and setting a 50% revenue split. The content is then reviewed by Grimes’ team for approval and then us to ensure it meets our DSP requirements.

It’s been exciting to see the results. Whilst in some cases the music being released only features Grimes’ voice, I think the more interesting music is the duets or where Grimes is a voice among other voices. That’s really cool.

What are the benefits of this model?

What we’re achieving with this is a couple of really great things. One is that Grimes has provided consent. Any music that’s coming through this technology has full clearance.

Then her team is receiving all the content to review prior to its distribution, so there’s control. And then there’s revenue share participation, so there’s compensation.

And finally, because we are asking for all of this to be documented correctly, we are creating transparency.

Every opportunity that we’re doing around AI really does have to pass those four principles to make sure that there is a responsible way to participate that is scalable.

As a distributor, what are the challenges you foresee when dealing with AI-generated music? For example, how do you deal with the huge scale of new music and obviously all the regulatory and administrative challenges?

I think the immediate fear is there’s going to be poor content flooding through, but that’s not the case. 

We still have to figure out how to scale the control side of it, but the goal here isn’t to have a bunch of low-quality content flooding the DSPs. And that’s why we put that control step in there. We want to make sure this is good content that the artist wants to get behind.

We want to make sure this is good content that the artist wants to get behind.

In terms of the regulatory steps, we’re watching this get figured out in real time. The White House had a press hearing just last week, and the week before the EU put out their stance on policy relating to AI.

We don’t know exactly how it’s going to play out, but if you are making sure that everything you’re engaging with is going to have the potential to create a more responsible ecosystem, then that’s how you can start participating early like we’re doing.

It’s important not to be afraid of new technology. I see AI as a great way for artists to be able to create better music. It’s not to replace human creativity, it’s how do we help enhance ourselves to do our jobs better? That’s the use case that I get very excited about.

And it’s opening up new ways for creators to be able to create in a way they couldn’t before. Cherie Hu from Water & Music just created a song with GrimesAI called “Eggroll”, and it’s really great. It’s amassed thousands of streams already.

The self-releasing DIY artist is the fastest growing segment of the industry. What is your take on the direction of travel for companies like TuneCore?

I joined TuneCore seven and a half years ago and the things that got me excited about the company continue to be the same, which is that there are no gatekeepers, and anyone can put out music. The artist is not worried about fitting into a box. They’re putting out what they feel and that tends to create more innovative music.

I think there are two big inflection points that happened simultaneously over the last few years to make self-releasing artists the fastest growing segment. Ultimately streaming has created more access to music around the world, so that has really grown the pool of listeners. And then platforms like TikTok and Shorts and Reels have enabled artists to have their music rise based on the merit of their music and the listener’s affinity for it.

At TuneCore we’re very leaned into, “How do we help artists grow? How do we do what labels historically used to do, but do it with technology?”

How we’re thinking about it is we are not just a distributor. We have moved into and have built programs that help drive artist development at scale in places where listening and discovery is happening.

Earlier this year we launched a program called TuneCore Accelerator where we work with all our DSP partners on their betas and opt our artists into those, and we are in the process of scaling those programs even further.

What are your goals for the future?

Everything’s moving more and more into digital and that’s where we’re innovating and really working with our partners to build not just discovery but development.

Everything’s moving more and more into digital and that’s where we’re innovating and really working with our partners to build not just discovery but development.

Our goal is to give artists a technology-based solution. And it’s up to the artist if they want to stay self-releasing and just hire a team around them, like Chance the Rapper or Russ, or move to a place where they can get more services. For example, TuneCore is a part of the Believe group and that’s where you can plug into more of a custom solution with teams around you to help with further development.

Ultimately, we’re a product company and we build technology that helps you get there. That’s where it gets very exciting in terms of what’s possible. But again, it’s up to the artists. There’s no one size fits all.

TuneCore recently published a study that surveyed 1,558 independent artist about AI and found that 50% were “aware and engaged in AI” with a “positive perception of its benefits and opportunities”. However, 39% were “unaware and apathetic toward AI” with “fears and concerns with the technology”.  You can read the full report here.