The Signal Report
Curated AI-music and industry news
German court rules Suno infringed copyright in landmark GEMA case.
The Munich Regional Court found that Suno broke German and U.S. copyright law by training its models on songs from GEMA's catalog and then reproducing them, in a case built around six tracks including "Rasputin" and "Forever Young." The ruling orders Suno to disclose revenue tied to the infringing use and pay damages still to be determined, and marks the first binding European court decision on whether unlicensed AI music training is legal. Suno says it disagrees with the ruling and is evaluating an appeal.
Read more →Majors and top indies agree on AI chart rules — IFPI adopts them the next day.
A coalition including Universal, Sony, Warner, Believe, BMG, Concord, HYBE, and several other independent labels proposed shared eligibility criteria for AI-assisted recordings appearing on official music charts, requiring any AI tools used to be properly licensed and the release to remain substantially human-made. IFPI adopted the framework the following day across the charts it directly runs — including its MENA, Southeast Asia, South Africa, and Latin American charts — and said it's working with national bodies including Australia's ARIA, France's SNEP, and South Korea's Circle Chart to extend it to 20+ more chart programs.
Read more →Study: the biggest AI "artists" have earned an estimated $6.1 million in streaming.
A new analysis from AI editing platform Kapwing estimates the ten highest-earning AI music acts have collectively pulled in around $6.1 million from more than 2.3 billion combined Spotify streams and YouTube views, based on estimated per-stream payout rates. Leading names include the 1950s-style vocalist Enlly Blue, nightcore/EDM creator Cherry Nightcore, and synth-rock project Masters of Prophecy. The underlying artist rankings draw partly on a lesser-known third-party database, so the topline figure is best read as a rough industry estimate rather than an audited number.
Read more →Spotify won't label AI music, so listeners are building their own detectors.
With Spotify still not tagging AI-generated tracks on its platform, third-party sites like SoullessMusic.com and SlopTracker.org have stepped in, letting users check whether an artist or track is likely AI-made. One case in point: R&B act Slime Dot, who has over 100,000 monthly Spotify listeners, denied being an AI creation in a magazine interview even as detection tools flagged the music as machine-generated. Spotify says it's relying on artist-submitted AI disclosures and profile-verification tools instead of platform-wide labeling for now.
Read more →Suno hit with two class-action suits over its data breach.
Following confirmation that a November 2025 security incident exposed personal data tied to 55.3 million Suno accounts, two separate class-action lawsuits were filed against the company in Massachusetts federal court. Both suits accuse Suno of inadequate data security and of failing to notify affected users for roughly nine months, until the breach became public through outside reporting. Suno has said the incident was "quickly contained" and that it does not store full payment card details.
Read more →Musicians' union expands its AI-licensing lawsuit against Universal and Warner.
The American Federation of Musicians filed an amended complaint against Universal Music Group and Warner Music Group, accusing both majors of breaching union contract terms by licensing members' recordings to AI companies Suno and Udio without compensation or credit. The amended filing names Warner Records and Atlantic Recording as defendants alongside UMG, with both majors now cleared to move to dismiss; opening briefs are due in mid-August. The case adds to a growing list of disputes over how AI licensing revenue gets shared with the musicians behind the underlying recordings.
Read more →Survey: US listeners are "more wary than willing" toward AI-made entertainment.
A new Luminate Intelligence report on audience attitudes found American consumers broadly skeptical about engaging with AI-generated music, film, and TV content, though reactions were mixed and some respondents were unsure whether AI had even been involved. The findings add a consumer-sentiment counterpoint to Luminate's earlier midyear data showing rising AI adoption among musicians themselves, pointing to a gap between how comfortable creators are using AI tools and how comfortable audiences are with the results.
Read more →AI-generated tracks now pass 50% of Deezer's daily uploads for the first time.
Deezer reported that fully AI-generated tracks hit a peak of roughly 90,000 per day in June, over half of all new music delivered to the platform — up from 44% in April and just 18% in early 2025. Despite the flood of uploads, Deezer says fully AI-generated tracks still account for only 1–3% of actual listening, with a large share of AI streams flagged as fraudulent. The platform will now start removing AI tracks tied to streaming fraud or that go unstreamed for six months or more.
Read more →Country hitmaker Joe Nichols: artists who say they never touch AI are "lying through their teeth."
Nashville songwriters built the demo for Nichols' new single "Say La V" partly by running a hummed melody through Suno, and Nichols then rebuilt and re-recorded the track more traditionally with his own band and a human horn section for the final release. In a separate interview, Nichols pushed back on artists who claim to avoid AI entirely, saying many are "lying through their teeth," while drawing his own line at using tools like Suno to test ideas rather than to write or perform for him.
Read more →A hack on Suno spills exactly which platforms it scraped for training data.
A hacker breached Suno and gave 404 Media source code detailing where its training data came from — YouTube Music, Deezer, Genius, and several stock-music and podcast libraries, amounting to what the report describes as decades of audio. The leak lines up with claims the RIAA has made in its ongoing copyright suit against Suno, which the company continues to defend as fair use. Suno confirmed the breach, saying it was contained quickly after being discovered last November.
Read more →BandLab Technologies acquires AI music studio Aiode.
BandLab picked up Aiode, an AI-powered music-creation platform built on audio-to-audio models developed alongside professional session musicians, giving BandLab a third production tool alongside its mobile app and Cakewalk's desktop software. Aiode lets creators direct performances section by section and generate alternate takes without touching the rest of a track. The deal folds a tool built on licensed audio from the start into a larger, more established music-tech portfolio.
Read more →Suno launches an iMessage extension for generating and sharing songs.
Suno added a keyboard extension that lets iPhone users create a song from a text or voice prompt directly inside a conversation, with recipients able to play it back and reply with songs of their own. The rollout landed the same week the hacked-training-data story broke — timing several outlets noted overshadowed what would otherwise have been a straightforward product launch.
Read more →Luminate's midyear report: streaming keeps climbing, AI still hasn't cracked the charts.
Luminate's 2026 Midyear Report found global on-demand audio streams up 9.8% and CD sales up 16% in the first half of the year, alongside growing acceptance of AI as a production tool — 54% of musicians surveyed said AI tools now have a place in their creative process. At the same time, the report noted that no fully AI-generated track had cracked the global Top 250, suggesting AI adoption as a tool is outpacing any breakthrough for AI-made songs as chart hits.
Read more →Suno responds to the industry's new AI-labeling push: "transparency is important."
Days after the RIAA/IFPI coalition unveiled its two-tier labeling plan, Suno issued a statement backing the broader push for transparency and pointing to its own investment in watermarking and audio fingerprinting tools that let artists disclose AI use — while stopping short of committing to the specific tags. The response comes as Suno remains in active litigation with Universal Music Group and Sony over its training data, with Warner the only major label it has settled with so far.
Read more →Music industry groups propose a two-tier system for labeling AI songs.
The RIAA and IFPI, joined by the Recording Academy, SAG-AFTRA, and several independent-label bodies, unveiled a plan for tagging AI music on streaming platforms — a bold "AI" badge for tracks generated entirely or primarily by AI, and a lowercase "ai" badge for human-led tracks that used AI for select elements. Whether major streaming services adopt the shared badges instead of their own existing systems is still an open question, since none has publicly committed yet.
Read more →A speed-altered clone of an indie band's own album took 94% of their Spotify royalties.
Philadelphia duo Makeshift Hammer discovered their catalog had been re-uploaded under a fake artist name, with tracks sped up or slowed down and paired with AI-generated cover art — a disguise good enough to slip past streaming fraud detection while racking up close to 50,000 listens per track. Because Spotify allocates royalties by stream share rather than a fixed pool per song, the clone diverted 94% of what the real recordings had earned, and a second clone targeted the band again about six months later.
Read more →Warner Music moves to dismiss musicians' union lawsuit over its AI licensing deals.
Warner asked a New York federal court to throw out the American Federation of Musicians' suit over its Suno and Udio licensing deals, arguing the labor agreement session musicians are suing under was signed in 2023 — before those platforms existed — and never contemplated payment for AI training uses. The union wants a cut of the licensing revenue and settlement money Warner and Universal collected; Warner says no contract entitles it to that.
Read more →LANDR sweetens its opt-in AI licensing program with cash advances.
LANDR expanded Fair Trade AI, its opt-in program for artists who let their catalogs be used to train AI tools, adding a $1 million fund for recoupable cash advances (open through July 20) and raising the artist revenue share from 20% to 25%. First payouts under the expanded terms go out later this month, with more than 30,000 artists already opted into the program.
Read more →Producer John Dahlbäck: AI for vocals and ideas, not full songs.
The Swedish house veteran — known for early collaborations with Avicii, Kaskade, and Benny Benassi — told MusicRadar he's gone from wary of AI to a "firm believer" in using it as a tool, mainly for vocal experiments and for sketching new ideas the same way he'd tweak a synth patch. He drew a clear line at the other end, saying he loses interest the moment someone uses AI to generate a finished song with no craft behind it.
Read more →The BBC lays out its first formal policy on AI in music.
BBC Director of Music Lorna Clarke said the broadcaster will prioritize music built on "meaningful human creativity," will still consider artists who use AI tools as part of their process, and wants artists and labels to disclose how AI was used in submitted tracks. The policy acknowledges it depends on an industry-wide labeling standard that didn't exist yet at the time — a gap the RIAA/IFPI proposal partly addresses.
Read more →Private equity firm CVC takes a majority stake in DistroKid.
CVC Capital Partners agreed to acquire a controlling interest in the independent distributor, which pushes a large share of the world's new music uploads to streaming, while longtime backer Insight Partners keeps a minority stake. The deal, expected to close in Q3 2026, puts one of the biggest upload pipelines for indie and AI-generated music under new ownership at a moment when private capital is pouring into music-adjacent infrastructure.
Read more →Suno is exploring a public developer API for the first time.
Suno's Chief Product Officer announced the company is investigating a self-serve API ahead of what he called a "partner powered model," and is inviting developers to apply for early access before any building starts. The move comes right after Suno closed a Series D above $5.4 billion, and would mark a shift from the unofficial API wrappers third parties have relied on so far to an official channel for building tools on top of AI-generated music.
Read more →Spotify makes its case for "licensed" AI remixes over open-ended generators.
Spotify's co-CEO publicly defended the company's paid AI remix feature, built on its Universal Music Group licensing deal, as a controlled alternative to tools like Suno and Udio. The add-on will let Premium subscribers generate covers and remixes of participating UMG artists' songs, with rights holders sharing in the resulting revenue — pricing and launch timing are still unannounced. It's a notable bet that "licensed and consent-based" can be a selling point in itself, not just a legal shield.
Read more →Australia's entire music industry went to Parliament to demand AI training rules.
A sweeping coalition of Australian music bodies — APRA AMCOS, ARIA, and more than a dozen others — delivered an open letter to the Federal Government calling for enforceable protections against unlicensed AI training, describing the situation as the largest intellectual property theft in the industry's history. The push came directly after The Atlantic's reporting that millions of Australian and New Zealand tracks appeared in AI training datasets without the knowledge or payment of the songwriters behind them.
Read the story →Suno faces a fresh lawsuit — this time from a production music library.
Belgium-based Jamendo filed copyright infringement claims against Suno on June 29, alleging its catalog of licensed tracks was used without authorization to train Suno's model. Jamendo's parent Winamp Group had filed similar claims against Nvidia the week prior. Suno — now valued at $5.4 billion — declined to comment and continues to argue its training practices fall under fair use, while still fighting separate lawsuits from Universal Music Group and Sony.
Read the story →TIDAL will demonetize fully AI-generated music starting July 15.
TIDAL announced that tracks identified as 100% AI-generated will be tagged with an "AI" badge, cut off from royalties and direct-to-fan sales, and that AI music impersonating existing artists will be removed outright. The policy is framed around protecting "organic creativity" — and it's the starkest platform-level stance yet from a major streaming service.
Read the story →AI-generated artists are now putting up real commercial numbers.
A new performance study tracking AI-generated acts across Spotify and YouTube found several synthetic artists pulling tens of millions of streams, with one breakout track alone estimated at over $600K in combined platform revenue. AI music has clearly moved past the novelty phase into legitimate commercial territory — at the same time major labels are still pursuing copyright litigation against the platforms behind the underlying models.
Read the report →Suno launched a grant program for indie artists — with a no-criticism clause in the fine print.
Suno introduced Spark this week, an incubator offering cash grants, marketing support, and writing-camp access to unsigned musicians who build music on the platform. The rollout landed alongside fresh backlash — including a public denunciation from SZA over the company's training data — and reporting that Spark's terms bar participating artists from publicly criticizing Suno during the program and after it ends. Worth watching for any AI-native artist sizing up what "platform support" actually costs in the fine print.
Read the story →Report: the same handful of music datasets keep turning up across AI labs.
The Atlantic identified four large collections of music tracks circulating among multiple AI developers, reviving questions about where the training data behind tools like Suno and Udio actually originates and how cleanly it's licensed. A useful reminder that "this tool credits its AI" and "this tool's training data is properly cleared" are two separate claims — a distinction SynRize tries to keep explicit by naming exactly which tools it uses, every time.
Read the report →This is currently hand-curated. Each item is summarized in our own words with a link to the original source — we don't republish full articles.