Music Recording

Sound Recording Industries (NAICS 5122) covers music recording, music publishing, and the broader recorded-music industry. The industry is highly concentrated at the major-label level: Universal Music Group (UMG, public on Euronext), Sony Music Entertainment (Sony Group subsidiary), and Warner Music Group (WMG, NASDAQ) collectively control over 70% of recorded-music revenue. Major independent labels include Concord, BMG, Believe, and Beggars Group. Music publishing (which holds song-writing rights separately from sound-recording rights) is similarly concentrated through Universal Music Publishing Group, Sony Music Publishing, and Warner Chappell. Combined global recorded-music revenue exceeds $30 billion annually after a decade of streaming-driven growth from the lows of the late-2000s CD-decline era. Streaming services (Spotify, Apple Music, YouTube Music, Amazon Music, Tidal) are now the primary revenue channel, having surpassed downloads, physical media, and synchronization combined.

AI is being applied across music creation, production, distribution, and rights management. The most controversial layer is generative music itself: Suno AI and Udio (Uncharted Labs) generate full songs from text prompts and have raised tens of millions in venture funding. Both face active copyright litigation from the major labels (UMG, Sony Music, Warner Music sued both in June 2024). Less controversial AI applications include production tools — iZotope, Splice, LANDR, and Output use ML for mixing, mastering, sample selection, and stem separation. Voice synthesis (ElevenLabs, Voicemod, Murf AI) is used for backing vocals and creator content. Music recommendation AI powers Spotify Discover Weekly, Apple Music's Stations, and YouTube Music. Rights-management AI tracks unauthorized use across YouTube (Content ID), TikTok, and Instagram.

The industry's structural question — whether AI-music distribution will run on licensed or litigated terms — gained its strongest licensed proof point with the Spotify and Universal Music agreement that lets Premium subscribers create AI-generated covers and remixes of UMG artists' songs, with revenue shared back to participating artists. The deal is framed around three principles — consent, credit, and compensation — and is the first major-label-licensed framework for consumer-facing fan-created AI music. It establishes a parallel path to the Suno and Udio cases: AI music sitting inside licensed economic terms rather than outside them, with the labels as economic partners rather than only as plaintiffs. Major labels are now simultaneously building AI products, licensing AI-music partners on the consumer surface, and litigating against generative-AI vendors that train on copyrighted catalog material without permission — three lanes at once, with the Spotify-UMG framework as the new template for the licensed lane.

The generative-music question is a referendum on training-data rights. If Suno and Udio prevail in their fair-use defenses, the economics of recorded music could be fundamentally altered; if they lose, the cost structure of generative-AI music platforms will rise sharply with licensing requirements. Producer and engineer roles are seeing meaningful AI augmentation — mastering and stem-separation AI accelerate routine work, though high-end production remains human-dominated. Session musician work has been declining for two decades and is further pressured by AI-instrument tools. Music supervision (placing songs in films, TV, and ads) is being AI-augmented through recommendation tools but the relationship-driven work remains insulated. Songwriting and composition remain protected by the 2023 WGA-style contractual framework in some major-label agreements. New roles are emerging: AI-music-rights coordinators, dataset-licensing specialists, and AI-voice-clone consent managers. Streaming-fraud detection (fake streams generated by bots) increasingly relies on AI to identify pattern signatures.