Music creators
Music publishers
About us
Main Menu
Payment and distribution
Self service
Register your music
FAQs
Music creators
Distribution categories
About your payment
Previous distributions
Payment and distribution
Television
Radio
Concerts
Online
Other areas
Payment and distribution
2024
2023
2022
Previous distributions
Music creators
Division of shares
Adapting and changing works
Self management of your works
Main Menu
Payment and distribution
Self-service
Works
Payment and distribution
Television
Radio
Concerts
Online
Other areas
Main Menu
Koda issues an official declaration expressly reserving their rightsholders’ rights in order to prevent AI services from extracting music and lyrics created by Koda's members.
iStock

Koda says no to AI- use without permission

Music creators’ content must not be used in the development of artificial intelligence (AI) without permission and payment. Therefore, Koda reserves the right to allow ChatGPT and other AI services to extract music and lyrics created by Koda’s members in order to train artificial intelligence.

Koda has now formally made a reservation against AI services being allowed to systematically collect material created by Koda's members to train their artificial intelligence. Doing so is part of Koda’s effort to ensure that Koda’s members are paid if their music and lyrics are to be included in the development of artificial intelligence.

"At Koda, we have an constant focus on safeguarding our members’ base conditions and protecting their rights. This is why we are now reserving our members’ rights in order to ensure that AI services do not collect our members’ music and lyrics without permission. To our minds, it is absolutely essential, obvious and natural that such services must compensate the music creators if they want to use their content to train their systems,’ says Kaare Struve, Director of Media licensing and strategy at Koda.

Any doubt must be eliminated


The step was taken due to the rules in section 11 b of the Danish Copyright Act, which is based on the copyright directive adopted in the EU in 2019 (also known as the text and data mining exception). A directive adopted back in the days before there was a real understanding of the copyright issues associated with artificial intelligence – and especially generative AI.

 

It is absolutely essential, obvious and natural that such services must compensate the music creators if they want to use their content - Kaare Struve

 

The law states that rightsholders must reserve their rights ‘in an appropriate manner’ in order to prevent the mining of data containing copyrighted content for use in training AI models. This differs significantly from how the copyright law normally works, where you have to ask for permission or enter into an agreement with the rightsholders before you can legally use protected content.

‘We firmly believe that it was never the intention of the directive to provide free access to using copyrighted content for the development and operation of AI systems – usually with profits in mind. That is why we are working both at EU level and back in Denmark to establish a clear legislative framework that will remove any doubts and remove the possibility of AI services and large tech companies systematically using the creative artists' works without fair payment,’ says Kaare Struve.