Believe and TuneCore Say Artists Must Give Consent Before Their Music Is Used for Suno AI Training

One of the biggest questions surrounding generative AI and music has always been simple: who gets to decide whether an artist's work is used to train an AI model?

Believe and its distribution platform TuneCore have now made their position clear.

Following Believe's licensing agreement with AI music company Suno, Believe founder and CEO Denis Ladegaillerie told Music Business Worldwide that artists must consent before their music can be included in AI training programs.

The company says it is using a two-stage consent process. Artists first choose whether they want their repertoire to be considered for AI-related opportunities at all. Those who opt in can then decide whether they wish to participate in individual partnerships such as the Suno agreement. Artists who decline the Suno opportunity will have their music excluded from that training process.

Believe describes its responsible-AI approach around four principles: consent, control, compensation and transparency.

AI licensing is becoming part of music rights management

For years, the debate around generative AI focused largely on whether copyrighted recordings had been used to build models without permission.

The industry is now beginning to enter another phase: licensed AI.

If AI platforms increasingly negotiate with labels, publishers, artists and distributors, musicians may start seeing new contracts and opt-in requests relating specifically to machine learning and generative products.

That creates an entirely new set of questions.

Which rights are being licensed?

For how long?

Can the artist withdraw?

Will the recording itself be used, or data extracted from it?

Can an AI-generated song imitate the artist's voice?

How will revenue be calculated?

And perhaps most importantly: who gets paid?

Artists should read before clicking “accept”

When new technologies arrive, contracts often evolve faster than artists' understanding of them.

That makes consent especially important.

Artists should avoid treating an AI opt-in box as just another setting inside a distributor dashboard.

It may involve valuable rights connected to recordings, compositions, performances, identity or future forms of usage that are still being developed.

Before participating, artists and labels should understand exactly what is being authorized and how compensation will work.

Believe's agreement does not settle the wider AI debate. Major disputes over copyright and AI training remain active across the music industry, and different companies are taking different approaches.

But the principle emerging here is important: artists should have meaningful control over whether their work participates in these systems.