Sony and UMG sue Suno over its label-backed v6 model
Two of the three majors are still at war with the AI music company that just signed up their rivals.
Sony Music Entertainment and Universal Music Group filed a new 45-page lawsuit against Suno in the US District Court for the District of Massachusetts, arguing that the company's licensed v6 models infringe their copyrights because they were trained on the outputs of the earlier, unlicensed ones. The labels' theory is blunt: Suno has acknowledged that v6 was built partly on the products of its previous models, and the filing argues that "training a 'new' model on the outputs of an infringing model does not eliminate the infringement; it launders it." Sony, UMG and several of their imprints are named as plaintiffs. Suno did not immediately comment.
The claim is a direct attack on the licensing-first playbook that the whole AI music sector has spent 2026 adopting. Suno's pitch for v6 — announced September 9 with Warner Music, BMG and Believe on board — was that the new family was trained on licensed catalogues and represented a clean break from the models the majors sued over in 2024. The labels' response is that a clean break is impossible if the new model learned from the old one: the derivative argument, if it holds, would reach every AI company that retrains a model on synthetic or legacy output rather than raw data. That is most of them.
It also sharpens the split inside the industry. Warner settled in November 2025 and helped build v6; BMG and Believe followed with deals. UMG and Sony are now the two holdouts litigating against a product their competitors already profit from — and their $9 billion damages claim against Suno is still live, as we covered when the platform passed 100 million users — Suno hits 100M users as UMG, Sony seek $9B.
A converted Cessna Caravan flew 3,199 miles across the United States with zero control inputs from its onboard safety pilot, Joby Aviation said Friday — the first fully autonomous coast-to-coast flight. The aircraft, which Joby calls the J208, handled its own taxiing, takeoffs, route navigation and landings while supervised remotely from Joby's autonomy headquarters in California and from Shaw Air Force Base in South Carolina, up to 2,323 miles away. It rerouted around thunderstorms in real time and slotted into Phoenix Deer Valley, one of the busiest general aviation airports in the country, without prior familiarity with any of the fields.
The eastbound leg finished near Kitty Hawk, North Carolina, with a low pass by the First Flight Monument, and the aircraft now flies home with stops in Raleigh, Washington, Louisville, Wichita, Oklahoma City, Salt Lake City and Portland. The strategic point is the retrofit: instead of certifying a new autonomous airframe from scratch, Joby is selling an autonomy stack that turns already-certified aircraft into uncrewed freight and medical-supply haulers — work that fits its growing defense business and the $17 million Air Force contract behind it.
What to watch: whether the labels' derivative-model theory survives a motion to dismiss — it will decide whether "we retrained on licensed data" is ever a clean defence.
A model trained on another model's output — is that a new product or a copy with extra steps? Tell us in the comments.
Sources: Variety · The Hollywood Reporter · Joby Aviation · TechCrunch
