Can AI music generators use Dutch singing voices?

In de media 18 March 2025 image: Possessed Photography via Unsplash

The voices of Dutch singers can be heard on music created by artificial intelligence (AI). Dirk Visser, Professor of Intellectual Property Law, spoke to Dutch public broadcaster NOS about copyright: ‘Your voice is personal data. It can’t just be used.'

Voices resembling those of Dutch singers including Jan Smit, Trijntje Oosterhuis and Herman van Veen can be heard on a number that was created by music generators Udio, Riffusion and Sonauto. These tools use AI to create music. With the emergence of music generators that are able to create music based on AI, the question arises whether these companies are breaching the copyright of the singers whose voices are used for the music.

According to Professor Visser, singers who want to prevent their voice from being used by an AI tool in such a recognisable way cannot invoke copyright law. After all, a piece of music is protected by copyright, but not the singer's voice itself. Artists can, however, invoke privacy laws: ‘Your voice is personal data. It can’t just be used.' That said, as yet no Dutch artist has taken an AI company to court, says the professor. ‘You have to have deep pockets and a lot of guts to take these kinds of tech companies to court.’