Suno Overhauls AI Music Models with Licensed Training Data Amid Copyright Legal Battles

AI music startup Suno launched a new family of models on Wednesday, dubbed Suno v6, marking a strategic pivot toward licensed training data in response to mounting copyright infringement lawsuits. The company says these models were developed using licensed content from music labels and distributors, including Warner Music Group, BMG, and Believe. Suno has been embroiled in multiple legal disputes with record labels accusing the startup of using copyrighted material to train its AI models, which generate music from user prompts. These lawsuits have escalated as the platform's popularity grows, prompting the company to shift its approach toward licensing partnerships. The tech-driven settlement with Warner Music Group last year was a pivotal moment, followed by a recent agreement with BMG. However, Suno continues to face legal challenges from other major players, including Sony and Universal Music Group, as well as high-profile artists like Jason Isbell. The company is also contending with class-action lawsuits from users, alleging various grievances related to the platform's operations. Suno emphasized that v6 models are not trained on the dataset used for previous model versions, signaling a clean break from the contentious past as it recalibrates its legal and ethical frameworks. This shift reflects a broader industry trend as AI companies seek to navigate the complex landscape of creative copyright. The startup introduced three distinct versions of the v6 model to cater to different user needs. The base Suno v6 model, available to paying subscribers, focuses on reliability and precise, controllable outputs. A more experimental variant, Suno v6 wild, also for premium users, encourages exploration and unpredictable results, ideal for ideation sessions. For broader access, Suno v6 mini offers a faster experience to all users. As the new lineup rolls out, Suno plans to retire its older models. The v6 models bring enhanced flexibility and features. Users can now edit specific song sections using prompts or lyric words, and even reference text, images, or videos as creative inputs for track generation. The platform also allows for instrument separation from samples, enabling users to craft new beats from existing audio. Looking ahead, Suno announced plans to introduce community-centric features, including a remixing capability. This will operate through a new opt-in program, where artists can choose to allow their songs to be used for AI-generated features in collaboration with music labels. Jack Brody, Suno's Chief Product Officer, highlighted the revenue-generating potential of derivative works for rights holders and artists. "The music ecosystem and our partners are always looking for ways to create more revenue opportunities," he explained in an interview with TechCrunch. "A significant part of this release is focused on establishing additional revenue streams." In parallel with the model rollout, Suno has taken proactive steps to address wider industry challenges. Following recent announcements, the company plans to watermark AI-generated songs and has introduced new download limits tied to account tiers to prevent misuse. On this front, Brody noted, "We can help manage broader ecosystem issues like streaming fraud and mass low-intent uploads to distributors. Ultimately, it's up to distributors and platforms to govern what content is allowed on their services." As the legal landscape evolves, Suno's shift to licensed training data and its new collaborative features represent a significant attempt to align with industry expectations. By fostering partnerships and introducing transparent practices, the company aims to bridge the gap between technological innovation and the protection of creative rights, setting a new course for AI music generation in 2026 and beyond.

via TechCrunch AI