Reservoir Media, one of the music industry's largest independent publishers, has filed copyright infringement lawsuits against AI music generator Suno and AI lab Anthropic, opening a new front in the escalating legal war over whether AI companies can train on copyrighted songs without a license.
The parallel complaints were reported Wednesday by Billboard and Music Business Worldwide. Against Suno, Reservoir alleges infringement covering more than 1,050 songs — including works associated with Joni Mitchell, Johnny Cash and Madonna — in what trade press reports describe as a copyright infringement and DMCA complaint seeking damages approaching $200 million. A separate lawsuit against Anthropic was filed the same day in the same court, according to Music Business Worldwide. For ongoing coverage of the industry's copyright battles, follow our breaking AI news.
Two Lawsuits, One Day
The structure of Reservoir's action is notable: rather than joining the existing label-led litigation against Suno, the independent publisher filed its own complaint, and paired it with a fresh case against Anthropic — a company whose music-related legal exposure until now has centered on song lyrics rather than recordings.
Billboard reported that the suits claim the AI companies used songs by Madonna and Johnny Cash, among others, to train their models, with the trade publication's headline quoting the suits' characterization that the systems were "meant to replace the artists."
Why the DMCA claims matter
Beyond straightforward copyright infringement, the reported inclusion of Digital Millennium Copyright Act claims is legally significant. DMCA theories let plaintiffs argue that AI companies removed or falsified copyright management information when ingesting recordings — and each violation can carry its own statutory damages, which is how the headline damages figure in the Suno case climbs toward the reported $200 million for roughly a thousand works.
The arithmetic explains why. Under US copyright law, statutory damages run up to $150,000 per work for willful infringement. Multiplied across the roughly 1,050 works Reservoir says are at issue, the theoretical ceiling before a jury exceeds $150 million — before DMCA violations are counted on top. Actual recoveries are typically far lower and most cases settle, but the exposure math gives plaintiffs like Reservoir enormous leverage in licensing negotiations that Suno would once have been able to defer.
It also clarifies what kind of company Reservoir is. The publicly traded independent publisher controls a deep catalog of recordings and publishing rights, and until now the AI music litigation has been dominated by the major labels and the major publishers. An independent filing on this scale suggests rights holders of every size now see litigation — not licensing-first negotiation — as the rational opening move against unlicensed training.
Suno, for its part, has built one of the most widely used consumer AI music services: a generator that produces complete songs, vocals and lyrics included, from simple text prompts. Its whole product, critics argue, is only possible because of the recordings it learned from — precisely the premise Reservoir's complaint attacks head-on.
Anthropic's Music Problem Keeps Growing
The new Anthropic complaint lands on the same day the company was in a US courtroom fighting music publishers over the fair-use defense underpinning its training practices. Anthropic has repeatedly argued that training on copyrighted text — including lyrics — is transformative fair use, a position publishers contest fiercely.
What Reservoir adds is a recording-catalog angle from the independent sector. The major publishers have been systematically litigating lyrics; an independent company now alleging unauthorized use of its recordings in training broadens the theory of harm and signals that independents intend to litigate separately rather than wait for the majors' cases to set precedent.
Suno's Legal Ledger Keeps Getting Longer
Suno, meanwhile, is juggling pressure on multiple fronts. Universal Music Group and Sony Music are already suing the company over what they have called model laundering — and Music Business Worldwide reported this week that, in response to that case, Suno has admitted it has not deleted all of the recordings used to train its AI. The company has also lost a landmark copyright case in Germany brought by collecting society GEMA.
For an AI music startup reportedly courting licensing deals and mainstream partnerships, an almost $200 million claim from a major independent publisher — backed by catalog names as recognizable as Mitchell, Cash and Madonna — is more than a legal nuisance. It raises the cost of every negotiation and hands licensors a template for their own demands.
What to Watch
Three threads will define the next phase. First, whether Reservoir's DMCA framing survives motions to dismiss — it is the piece of the complaint with the widest financial consequence. Second, whether the Anthropic case gets consolidated or coordinated with the publishers' existing lyrics litigation, which would put recordings and lyrics before the same judge. And third, whether other independent publishers and labels follow Reservoir's lead, converting what has been a majors-versus-startups fight into an industry-wide reckoning over training data.
Either way, the message from Wednesday's filings is hard to miss: the people who own the songs are no longer asking AI companies to license them — they are daring them to defend building without asking.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
