Typing an AI prompt is not 'active' music creation (www.theverge.com)

AI music startup Suno just closed a $250 million funding round at a reported $2.45 billion valuation even as it faces lawsuits from the “big three” labels, the RIAA and some indie artists for allegedly training its models on copyrighted recordings. CEO Mikey Shulman’s comment that “typing an AI prompt is not ‘really active’ music creation” has provoked backlash from musicians and critics who say Suno’s push-button Create feature—text-prompt generation of full tracks—actually sidelines human craft. Suno also offers Studio, a more DAW-like tool that can transform audio (hum-to-instrument), separate stems, record live takes and generate drums or vocals from its model, but access requires a Premier subscription (~$24/mo), leaving questions about whether it truly democratizes production versus replacing skill-building. The dispute matters technically and economically: at stake are copyright and training-data practices, the meaning of authorship when models output complete songs, and the market effects of near-instant mass music generation. Platforms like Deezer, Qobuz and Spotify are already down-ranking or removing fully AI-generated tracks, reflecting worries that limitless, low-effort output undermines scarcity and perceived value. For creators and ML practitioners this raises core issues—data provenance, licensing, model design that preserves attribution and control, and the downstream cultural impact of commoditizing musical creation.