After a judge refused to let Sony bolt 30,000-plus songs onto its existing case, the label filed a fresh copyright suit against the AI music generator in Manhattan — leaving Sony the last of the three majors still fighting Udio in court while Universal and Warner have moved on to licensing deals.
2026-07-21 · by Moe Ameen
On July 20, 2026, Sony Music Entertainment and nine affiliated labels filed a new copyright-infringement lawsuit against Udio in the U.S. District Court for the Southern District of New York (Manhattan). The complaint asserts that Udio copied 30,117 sound recordings — spanning artists from Beyoncé and Michael Jackson to Bob Dylan and Dolly Parton — without a license to train its generative music models. Sony is seeking statutory damages of up to $150,000 per infringed work, plus up to $2,500 for each act of DMCA circumvention, and a jury trial. At that per-work maximum, 30,117 recordings put Udio's theoretical damages exposure as high as roughly $4.5 billion.
The second suit is a direct consequence of a procedural ruling. Sony's original case — part of the RIAA-coordinated actions filed in June 2024 on behalf of the major labels — named 333 songs, an exposure on the order of $50 million. When Sony moved to add the 30,117 additional recordings to that existing case, Judge Alvin K. Hellerstein denied the motion in a June 29, 2026 ruling, finding that folding tens of thousands of new works in so late would substantially prejudice Udio and unduly delay the case. The court acknowledged Sony's right to pursue those works — just not inside the first lawsuit. Sony's answer was to file them as a separate action.
Sony is now the only one of the three major labels still litigating against Udio. Universal Music Group and Warner Music Group both dropped their claims in late 2025 after reaching settlements and licensing deals, and Udio has since signed additional agreements with Merlin, Kobalt, Believe, and the National Music Publishers' Association. Those deals reportedly require Udio to build a new model trained on pre-cleared data and to keep AI-generated music inside a "walled garden" on its own platform. Sony's complaint frames Udio's pivot pointedly: "Udio's belated embrace of licensing only underscores the unlawfulness of its decision to copy Plaintiffs' copyrighted sound recordings, without a license, in the first place." Udio has previously argued that training on publicly available audio is fair use. The core question — whether training AI models on copyrighted recordings without permission is infringement or fair use — remains unsettled by any final ruling.
The Sony–Udio fight is a reminder that the soundtrack is the contested layer, not the content operation around it. Let the labels, platforms, and generators sort out who owns the audio; the part you actually control — the video, the captions, the visual brand, the distribution — is where a durable content business lives, and that is the layer Kompozy runs. Kompozy doesn't generate music, and it won't pretend an unsettled fair-use argument is safe to build a monetized campaign on. Instead it lets you keep the soundtrack decision deliberate and swappable — a track you're licensed to use, a platform's native licensed audio library, or a licensed video-to-music tool — while the finished piece gets built and shipped around it.
Concretely, if you make music-driven short-form, Kompozy is where the video is produced and distributed. It generates Clipped Shorts, Listicle Video, and Naturalistic Video, composites a face-locked persona into brand-exact HyperFrames templates, burns in captions, and reframes to 9:16, 1:1, and 16:9 — all governed by the Persona Brief so a week of posts holds one voice. From a single idea it fans out 25–35 outputs and schedules them across nine social platforms plus blog and email, with Autopilot and a per-post review queue handling the publishing. Because every visual input is something you own or license — brand assets, Pexels b-roll, gpt-image and Gemini visuals, your persona's face — the walled-garden and licensing turbulence over AI audio never puts your publishing pipeline on hold. You choose the music source; Kompozy makes everything else and ships it.
Sony tried to add 30,117 recordings to its original 2024 case, but on June 29, 2026 Judge Alvin K. Hellerstein denied the motion, finding that adding tens of thousands of works so late would substantially prejudice Udio and unduly delay the case. The court acknowledged Sony's right to pursue those works — just not inside the first lawsuit — so Sony filed them as a separate action on July 20, 2026 in the Southern District of New York.
Sony is seeking statutory damages of up to $150,000 per infringed work plus up to $2,500 for each act of DMCA circumvention. Across 30,117 asserted recordings, that puts theoretical exposure as high as about $4.5 billion — a jump from the roughly $50 million tied to the 333 songs in the original case. These are the maximums a plaintiff can seek, not an amount any court has awarded.
The legal question is unsettled. Udio has signed licensing deals with Universal, Warner, Merlin, Kobalt, Believe, and the NMPA, but Sony is still litigating, and the broader question of whether training on copyrighted recordings without a license is fair use has not been finally decided. For monetized or client work, the safest path is music you can prove you're licensed to use; consult the current terms and, for high-stakes campaigns, a lawyer.
The contested layer is the audio, not the video around it. You can keep producing captioned, on-brand short-form video in a content engine like Kompozy — Clipped Shorts, Listicle Video, Naturalistic Video, and persona formats — while choosing a soundtrack you can stand behind: a licensed track, platform-native licensed audio, or a licensed video-to-music tool. That keeps your publishing pipeline independent of any single generator's legal exposure.