Regulation

Sony Music Escalates Copyright Fight With Second Udio Suit

mm
Add Unite.AI to your preferred sources on Google

Sony Music Entertainment has opened a second front in its copyright fight with Udio, filing a new lawsuit on July 20, 2026 that accuses the AI music company of copying 30,117 of its sound recordings to train generative models without a license. Brought in federal court in Manhattan alongside nine affiliated Sony labels, the complaint pushes Udio’s potential damages from roughly $50 million to as much as $4.5 billion, and sharpens the question hanging over the whole AI industry: whether training a model on copyrighted work without permission is fair use.

Sony is now the only one of the three major labels still fighting Udio in court. Universal Music Group and Warner Music (WMG ) Group each dropped their claims after settling and signing licensing deals with the startup in late 2025. Sony’s choice to escalate rather than settle turns the case into a test of whether litigation still carries leverage once most of the market has moved to licensing.

Turning rivals’ deals into evidence

Sony’s complaint takes direct aim at the fair-use defense, the doctrine that permits unlicensed use of protected work for sufficiently transformative purposes and the argument Udio has leaned on to justify feeding recordings into its models. Sony’s counter targets one of the four factors courts weigh in a fair-use analysis: whether the use harms an actual or potential market for the original work.

Here Sony turns its rivals’ settlements into ammunition. Because Udio has now signed training licenses with Universal, Warner, Kobalt, Merlin, Believe and the National Music Publishers’ Association, Sony argues that a licensing market for AI training data plainly exists, and that Udio’s own payments prove it. “A company that pays to license the very inputs at issue cannot credibly maintain that no market exists for those inputs,” the complaint states. The industry-wide pivot from suing to licensing, which had looked like AI’s truce with the music business, becomes the labels’ sharpest weapon against the one company still in court.

Sony is careful to cast itself as pro-licensing rather than anti-AI, pointing to its own AI partnerships with Spotify and the startup Klay and arguing that machines and human artists can coexist through ordinary market licensing. The subtext is a demand: pay for the inputs, and there is a deal to be had.

Why Sony had to start over

The scale of the new claim is a product of discovery. Sony’s original 2024 suit, coordinated by the Recording Industry Association of America on behalf of all three majors, named just 333 recordings, which the label now calls a “minuscule, illustrative fraction” of what Udio actually copied. After discovery gave Sony access to Udio’s training data, it says audio fingerprinting matched hundreds of thousands of its recordings; the 30,117 in the new suit are a subset. Many, the complaint alleges, were stream-ripped from YouTube with a tool that stripped the platform’s anti-copying protections.

Sony first tried to fold those recordings into its existing case, but the court refused. In a June 29, 2026 ruling, Judge Alvin K. Hellerstein found that adding tens of thousands of works so late would prejudice Udio and delay the case, while spelling out the exit Sony has now taken. “I recognize that plaintiffs have the right to seek to stop infringement of, and recover damages for, all copyrighted works,” he wrote. “But there is no requirement that it be done in this lawsuit.”

The new complaint brings three claims:

  • copying recordings made in or after 1972, covered by standard federal copyright;
  • copying older recordings, brought under federal protection by a 2018 law covering pre-1972 sound recordings; and
  • stripping YouTube’s technical safeguards, a violation of federal anti-circumvention rules.

Sony is seeking the statutory maximum of $150,000 per recording, plus up to $2,500 for each act of circumvention, and a court order halting the alleged infringement — the arithmetic behind the $4.5 billion figure.

A fight the rest of the industry has left

Udio, built by Uncharted Labs and founded by former Google DeepMind researchers, launched in April 2024 and has spent the past year converting adversaries into partners. Under the deals it struck with Universal and Warner, it agreed to build a fresh model trained only on cleared music and to keep AI-generated songs inside a “walled garden” on its platform. In answering the original suit, it acknowledged its models were built by showing them “a vast amount” of recordings that “presumably” included the labels’, and it has admitted pulling audio from YouTube, while maintaining the training is fair use.

That defense remains legally untested, and it is the pivotal unresolved question in dozens of AI copyright suits. The closest analog to reach any conclusion, Anthropic’s settlement over pirated books, sidestepped it: a judge found the training itself transformative but the underlying piracy unlawful, and the company paid to settle rather than win a precedent.

The same fight is running against Udio’s rival Suno, where Universal and Sony are pressing parallel claims in Massachusetts and seeking to add 61,026 recordings, more than double the number now at issue against Udio. Whatever the Manhattan court decides about Sony’s market-harm theory will not stay contained to music; it feeds directly into how AI companies everywhere are expected to pay for training data, a question steadily reshaping copyright law.

Sophie Denar is an AI-generated journalist at Unite.AI, covering artificial intelligence policy, regulation, and governance across global markets. Her work focuses on how national and international regulatory frameworks shape the development, deployment, and commercialization of AI technologies over the long term.

With a diplomatic and globally informed perspective, Sophie tracks policy initiatives from governments, multilateral institutions, and standards bodies, analyzing how differing regulatory approaches affect innovation, competition, and market access. She pays particular attention to cross-border implications, compliance challenges, and the balance between risk management and technological progress.

Articles authored by Sophie Denar are AI-generated and reviewed by Unite.AI’s editorial team to ensure accuracy, neutrality, and responsible coverage of AI policy and regulatory developments worldwide.