Streaming giants have signaled support for the recorded music industry’s new AI labeling initiative, yet they caution that broken metadata pipelines could render the two proposed tags ineffective. The Digital Media Association (DIMA), representing Spotify, Apple Music, YouTube, Amazon, and TIDAL, issued a response last week to the announcement by the RIAA, IFPI, and a coalition of creator groups including SAG-AFTRA and the Recording Academy. While DIMA President Graham Davies stated the group is “following today’s announcement closely,” he emphasized that the system only works if accurate AI metadata travels seamlessly from creator to fan.
Industry Push for Transparency Tags
The proposed system introduces two distinct tags to distinguish between “AI-generated” and “AI-assisted” music, functioning similarly to current “explicit” markers on streaming platforms. The “AI-generated” tag applies to tracks built entirely from a text prompt or where machines produced lead vocals and main instruments, while “AI-assisted” flags tracks where humans crafted the piece but AI contributed to expressive aspects. RIAA Chairman Mitch Glazier argued that listeners are willing to embrace AI music if they know when a real person was involved, calling transparency the best way to balance flexibility with trust.
Crucially, the tags currently do not specify how AI was used in lyrics, composition, or visual art, nor do they cover AI’s role in cover art or music videos. The Wall Street Journal reported that AI usage is flagged voluntarily by artists, labels, and distributors, meaning the system still relies on an honor system rather than automated detection.
Metadata Breakdowns and Platform Responses
Despite the industry’s push, streamers warn that metadata failures could undermine the program. DIMA noted its members rely on industry partners and standards bodies like DDEX to build a robust supply chain for consumer trust. Spotify announced support for the newly proposed DDEX standard for AI disclosures in music credits and began testing AI tags in song credits in April, though only those specified by artists or labels. Apple Music launched a similar system in March that also depends on label disclosure rather than platform-level detection.
Deezer, which already detects and tags AI music at the platform level, reported receiving 75,000 fully AI-generated tracks daily in April—over 44% of new uploads—and found up to 85% of streams on such tracks were fraudulent. TIDAL and Qobuz have also implemented policies to tag fully AI-generated tracks and remove those impersonating artists or faking streams. The success of the new labeling program hinges on whether labels and distributors can deliver the accurate metadata streamers demand.
For editorial consideration and industry coverage inquiries, contact Radio Facts.
