Photo by Aleksandar Pasaric on Pexels

Facing fierce opposition from its user base, SoundCloud has amended its terms of service to better protect artists’ content from unauthorized AI training. The initial update, implemented in February of last year, sparked concerns that uploaded music could be used to train AI models without explicit consent. CEO Eliah Seton admitted the previous language was overly ambiguous, leading to widespread apprehension within the artist community.

SoundCloud now unequivocally states that it has not, and will not, use artist content to train AI models without prior authorization. The revised terms emphasize consent, transparency, and artist control, specifically prohibiting the use of content to train generative AI models that replicate an artist’s voice, music, or likeness without their express permission.

Despite these changes, some skepticism persists. Tech ethicist Ed Newton-Rex suggests that the revised language might still permit the training of AI models that, while not directly replicating an artist’s style, could still create music that competes with them. The debate continues regarding the optimal balance between technological advancement and safeguarding artists’ rights in the rapidly evolving landscape of AI-generated music.