Hackers published code and dataset indexes this week that suggest Suno scraped millions of hours of music, lyrics, and podcasts from services such as YouTube, Deezer, and Genius MusicTech. Reporting and industry complaints say major labels and the RIAA are using discovery to tie Suno’s training data to copyrighted tracks, intensifying legal scrutiny The Verge. Those contested disclosures set the tone for litigation, policy, and business moves.
Suno scraping claims tested as hacked code surfaces
Hackers posted code and dataset indexes that, according to the exposé, indicate Suno ingested vast catalogs and lyric corpora from streaming and lyric sites, which would expand the material available to its generative models MusicTech. The fragments copy an image of large-scale crawling and raise immediate technical questions about what was collected, how it was normalized for training, and whether common platform mirrors or API endpoints were used to assemble those sets.
Separate reporting and industry complaints show major labels and trade groups are already using discovery to link model inputs back to copyrighted tracks, and plaintiffs argue those practices can amount to unauthorized copying at scale The Verge. Suno’s public defenses about originality and filtering now face sharper legal scrutiny, making the provenance and processing of training data both a technical and a courtroom issue.
Sony presses litigation route against Udio amid procedural limits
Sony Music filed a fresh copyright suit against AI platform Udio, alleging its models were trained on more than 30,000 sound recordings and that those files were used without permission Music Business Worldwide. The new filing follows a judicial decision that limited Sony’s earlier effort to expand its complaint, so this case reads as a recalibrated legal strategy rather than a simple replay of prior claims.
Beyond the headline numbers, the suit underscores a labels-versus-platform fight over discovery, disclosure, and the degree to which courts will compel transparency about training sets and model inputs Music Business Worldwide. For lawyers and product teams alike, procedural rulings in this action will shape negotiation leverage and commercial settlement dynamics as much as any legal doctrine.
Rights groups and distribution partners push licensing and tracking options
Australia’s collecting society APRA AMCOS publicly welcomed Prime Minister Albanese’s backing for creators and the new Office of AI, which the government intends to use to design licensing frameworks that require consent and payment for use of creative works CISAC. APRA AMCOS also emphasized the need to protect indigenous cultural intellectual property in those conversations, signaling that national policy will try to balance cultural rights with AI development.
On the commercial front, Sureel AI’s partnership with Symphonic Distribution creates a revenue-tracking and licensing pathway aimed at independent artists who want transparent payment when AI models use their compositions Warner Music Group. Those tools offer a complementary path to litigation by making it technically feasible to trace usage and route payments, which could lower friction for creators seeking practical control over AI-driven exploitation.
