AI Copyright: Training-Data Lawsuits and Licences — 2026-09-26
This week's big story is the collapse-and-licensing dynamic in AI music: labels that sued Suno and Udio are now licensing their catalogs to the same platforms, while the payouts remain undisclosed and musicians demand answers. In Germany, the Munich Regional Court ruled that Suno's memorisation of musical works constitutes copyright infringement in the GEMA case. Meanwhile, US motion practice in NYT v. OpenAI continues to frame the fair-use question that could reshape generative AI economics.
AI Copyright: Training-Data Lawsuits and Licences — 2026-09-26
Top developments
Labels turn from plaintiffs to licensors in AI music
As of this week, record labels that accused AI music startups of copyright infringement are now licensing their catalogs to the very same platforms — but payouts to artists remain undisclosed and lawsuits continue, leaving musicians asking who actually benefits. The reporting highlights a structural tension: licensing revenue flows to labels, not necessarily to the artists whose recordings underpin the deals.
Universal, Sony and DistroKid remain in court over AI music
New September 2026 coverage details that Universal, Sony and DistroKid are back in litigation over AI music, even as several platforms have already settled. The piece frames what the ongoing suits mean for working musicians' livelihoods — a reminder that licensing deals have not extinguished the litigation docket.
Munich court rules Suno's memorisation of music infringes copyright
The LG München I held that memorised musical works and similar AI outputs from Suno constitute copyright infringements in the GEMA case. The judgment is not yet final (nicht rechtskräftig), and the German-language legal press flags it as a significant test of the memorisation theory in the EU.
US DOJ position on AI training ripples internationally
Japanese-language tech media this week examined why the US Justice Department's view — that barring AI training on copyrighted material could "severely hamper" tech progress — poses risks for creators, noting that OpenAI, Anthropic and peers have depended on vast copyrighted knowledge bases for model training since generative AI's emergence in early 2023.
Local view
- Germany: Datenschutzticker reports the Munich Regional Court's GEMA v. Suno decision on memorisation, stressing that the verdict is not legally final and creates pressure for licensing frameworks in Europe.
- Japan: ZDNET Japan frames the US DOJ's fair-use leaning as a potential threat to copyright holders, reflecting local anxiety now that Japanese newspaper groups have also sued US AI firms over training data.
Context & numbers
- OpenAI has roughly two dozen publicly announced publisher/data deals, far ahead of Microsoft and Meta in volume; key figures include the reported $250M over five years with News Corp (the largest disclosed deal) and Apple's reported multi-year news-archive offers of at least $50M.
- Licensing-deal volume trajectory: 0 in 2023, growing to a projected 36 deals in 2026 — the market is becoming infrastructure, not experimentation.
- The Bartz v. Anthropic settlement requires $3,000 per pirated book used to train its chatbot, part of the $1.5B package; authors are now wrangling with publishers over who collects the funds.
On the radar
- The finality (appeal) question in GEMA v. Suno at LG München I — if upheld, memorisation becomes a stronger enforcement hook across the EU.
- Undisclosed terms of the label–AI music licensing deals: artist groups are pushing for transparency on who benefits; expect pressure over payout structures.
- CDT Europe's September AI Bulletin flags implementation milestones under the EU AI Act and the Digital Omnibus process in Brussels this autumn — with potential knock-on effects for TDM exceptions.
This content was collected, curated, and summarized entirely by AI — including how and what to gather. It may contain inaccuracies. Crew does not guarantee the accuracy of any information presented here. Always verify facts on your own before acting on them. Crew assumes no legal liability for any consequences arising from reliance on this content.
Powered by
