The music industry unveils a two-tier AI labeling system for streaming, TIDAL becomes the first platform to cut royalties on AI-generated tracks, and two landmark court rulings on AI music training loom in Boston and Munich.

Music industry coalition launches AI labeling standard for streaming platforms

On July 10, the RIAA, IFPI, Recording Academy, SAG-AFTRA, A2IM, WIN, and IMPALA jointly announced a voluntary labeling program distinguishing "AI-Generated" from "AI-Assisted" recordings on streaming services. The system uses two square icons — "AI" in all caps for fully generative tracks, lowercase "ai" for human-led recordings that use AI for some expressive elements. To qualify as merely "AI-Assisted," humans must have performed the lead vocal and primary instruments.

The scope is deliberately narrow: it covers sound recordings only, not lyrics, composition, videos, or cover art. The labels are voluntary, applied by artists and distributors — no enforcement mechanism exists yet. Think of it as the "Explicit" tag for the generative era. If you distribute AI-assisted music, start thinking about how you'd classify your workflow now, because this framework will likely inform whatever mandatory requirements follow the EU AI Act's Article 50 transparency rules taking effect August 2.

Source: Variety, Digital Music News, Music Ally, Hollywood Reporter

TIDAL will stop paying royalties on fully AI-generated music starting July 15

TIDAL announced that tracks identified as wholly AI-generated will lose royalty eligibility effective July 15. An external detection partner will flag tracks where every component was made with generative AI. Those tracks stay streamable — they just get an "AI" badge and lose access to royalty payments, tipping, and direct fan sales. Tracks involving fraudulent activity (impersonating artists, deceiving listeners) get removed entirely.

This is the first major streaming platform to formally demonetize AI-only music rather than simply removing it. The policy is a "living document" that TIDAL expects to iterate on as detection and evasion evolve. For producers using AI as one tool among many, the distinction matters: hybrid work keeps its royalties. For anyone running prompt-to-release pipelines, TIDAL just became a distribution dead end.

Source: TechCrunch, Variety, MusicNews.com

Two landmark AI music training rulings due this month in Boston and Munich

As reported July 10, federal courts on two continents are set to deliver rulings that could define the legal ground rules for AI music training. In Boston, Chief Judge Saylor will hear Sony's summary-judgment motion in Sony v. Suno — the first time a U.S. federal judge will directly address whether training a music AI on copyrighted recordings qualifies as fair use. In Munich, the Regional Court is expected to rule on GEMA v. Suno by July 31, testing the EU's text-and-data-mining exception against a collecting society's rights.

The stakes are asymmetric: Sony is the last major label that declined to settle (Warner settled with Suno in November 2025, Universal with Udio earlier). Audio fingerprinting in discovery revealed millions of copyrighted recordings in Suno's training data. If the court finds training is not fair use, expect every AI music company to face renegotiation. If it's ruled transformative, the labels' negotiating position collapses. Either way, the music AI business model gets repriced this month.

Source: TechTimes, Music Times, Bloomberg Law

Boris FX acquires iZotope from Native Instruments' wreckage

Boris FX announced July 2 that it has acquired iZotope — maker of RX, Ozone, Neutron, and Nectar — pulling it out of the Native Instruments family as part of the fallout from NI's January insolvency. The deal closed alongside inMusic's acquisition of Native Instruments proper on June 30. iZotope will operate as a dedicated unit within Boris FX with its engineering and product teams intact.

For audio practitioners, the immediate signal is stability: licenses and subscriptions carry over unchanged. The longer-term question is whether Boris FX's video post-production DNA accelerates iZotope's AI-powered audio tools (RX's dialogue isolation, Ozone's Master Assistant) toward tighter video-audio integration. Boris FX founder Boris Yamnitsky has historically been a product-focused owner — a better home than a distressed conglomerate.

Source: Synth Anatomy, CDM, Bedroom Producers Blog, Post Magazine

Suno opens developer API partner program at $5.4B valuation

Suno CPO Jack Brody posted an intake form on July 1 inviting developers to apply for early access to an official Suno API — the company's first move toward letting third-party apps embed music generation directly. Brody described this as a "partner-powered model," starting with a curated group focused on "experiences generative music makes possible for the first time." No public timeline was given.

This arrives fresh off Suno's $400M Series D at a $5.4B valuation in June. Until now, the only developer access has been through unofficial wrappers of dubious legality. An official API would unlock adaptive music in games, personalized audio in apps, and dynamic soundtrack generation — but the copyright uncertainty from the Sony case makes any commercial integration a calculated bet until the court rules.

Source: Music Business Worldwide, Digital Music News

ElevenLabs ships production tools for ElevenMusic: Voice to Song, Loop Studio, and more

ElevenLabs expanded its music platform on July 2 with four new ElevenMusic Tools aimed at different production stages. Voice to Song takes a rough phone-recorded vocal and outputs a studio-quality track preserving your melody and lyrics. Loop Studio generates clean instrumental loops from a genre pick and BPM. Additional tools handle genre-flipping and acoustic stripping of existing tracks. All run through the same interface as ElevenLabs' voice and SFX tools.

The Voice to Song tool is the most interesting addition — it solves the specific problem of capturing a melodic idea without needing an instrument or production setup. Record a voice memo, get back a produced track. Combined with their Inpainting API for surgical edits to generated audio, ElevenLabs is building a credible end-to-end music production pipeline for non-instrumentalists.

Source: ElevenLabs Blog, Promethean AI, Releasebot
Tool of the week
ElevenLabs Voice to Song

Part of the new ElevenMusic Tools suite, Voice to Song turns a rough vocal recording — hummed melody, sung idea, phone voice memo — into a fully produced track that preserves your original melody and lyrics. It's not generating music from a text prompt; it's completing production around your existing musical idea. That distinction matters for anyone who writes melodies but doesn't produce.

The output quality reflects ElevenLabs' audio model strength — coherent arrangements, genre-appropriate instrumentation, broadcast-ready fidelity. Combined with Loop Studio for rhythm creation and genre-flip tools for exploring variations, it's a meaningful workflow for songwriters who want to hear ideas realized without opening a DAW. The limiting factor is creative control: you guide with your vocal, but arrangement decisions are the model's.

Included with ElevenLabs subscription (from $5/mo) · Web · elevenlabs.io/music
Worth watching
The emerging two-track system: labeled and demonetized

This week's RIAA labeling announcement and TIDAL's royalty ban are two data points forming a clear trajectory. The industry is building a system where AI-generated music isn't banned from platforms — it's tagged and economically marginalized. TIDAL's approach (keep streaming, kill revenue) paired with the coalition's labeling standard (make AI usage visible to listeners) creates a world where AI music exists in a clearly marked, royalty-free tier alongside human-created work that earns.

The open question is enforcement. Voluntary tagging has obvious evasion problems — why would a bad actor self-label? TIDAL's reliance on an external detection partner suggests algorithmic classification will be the real gatekeeper. Watch for Spotify and Apple Music to announce their own policies before the EU AI Act deadline on August 2. The distinction between "AI-generated" and "AI-assisted" will be the most contested boundary in music distribution for the rest of this year.