<\/p>\n
Apple Music<\/a> is rolling out what it calls \u201cTransparency Tags,\u201d a system for flagging AI-generated content on its platform. Before celebrating the dawn of radical honesty in streaming, though, there\u2019s a catch: the system appears to rely largely on record labels choosing to actually use it.<\/p>\n On Wednesday (4 March), Apple<\/a> sent a newsletter to industry partners announcing that AI disclosures would now be a \u201cdelivery requirement\u201d for content submitted to the service.<\/p>\n The tags cover four categories: Artwork, Track, Composition, and Music Video, each intended<\/i> to indicate when AI contributed a \u201cmaterial portion\u201d of the work.<\/p>\n \u201cProper tagging of content is the first step in giving the music industry the data and tools needed to develop thoughtful policies around AI,\u201d the newsletter states. \u201cWe believe labels and distributors must take an active role in reporting when the content they deliver is created using AI.\u201d<\/p>\n \u201cThese new tagging requirements provide a concrete first step toward the transparency necessary for the industry to establish best practices and policies that work for everyone.\u201d Apple\u2019s technical specification<\/a> describes the tags as \u201coptional\u201d \u2013 at least for now \u2013 and the system does not appear to include any visible enforcement mechanism or verification process. \u201cIf omitted, none is assumed,\u201d the notes state.<\/p>\n In practice, that likely means labels can tag AI-generated elements \u2013 be it a drum loop, lyric line, or album artwork \u2013 if they choose to disclose it. If they don\u2019t, nothing changes.<\/p>\n Given the sheer scale of AI-generated uploads, that limitation could prove significant. Last September, Spotify<\/a> introduced similar AI disclosure labels, alongside a policy allowing the removal of tracks with unauthorised AI generated voice clones<\/a>.<\/p>\n Other platforms have also taken a more proactive approach. Deezer<\/a>, for once, implemented an automated AI-detection system<\/a> more than a year ago. The company says it now receives over 60,000 AI-generated songs every day, and its detection tools have identified more than 13.4 million AI-created tracks on the service.<\/p>\n Nevertheless, Apple\u2019s Transparency Tags represent a step toward clearer disclosure \u2013 though relying on self-reporting alone is unlikely to slow the flood of AI-generated music.<\/p>\n<\/div>\n Get the latest news, reviews and tutorials to your inbox.<\/p>\n<\/div>\n<\/div>\n
Exactly how that transparency will be enforced remains unclear.<\/p>\nGet the MusicTech newsletter<\/h2>\n