Sony Music says it had asked streaming platforms to remove more than 260,000 fake tracks by the end of September—nearly twice the 135,000 it had flagged six months earlier. The company described a surge of generative-AI recordings that imitate artists’ voices and likenesses without permission, including soundalikes associated with Adele, Harry Styles, Britney Spears, Queen and Michael Jackson.
The number matters because it turns a debate about hypothetical misuse into an industrial-scale rights problem. A convincing synthetic performance can now be made, titled, distributed and surfaced beside authentic work without the artist entering a studio. The harm is not limited to copyright. A false track can trade on a performer’s identity, confuse listeners, weaken search results and attach an artist’s name to material they never approved.
Two problems are colliding
AI music is not one category. Some musicians use generative tools as part of a clearly disclosed, human-led process. Other releases impersonate known performers. A third group consists of tracks generated in bulk and paired with automated plays to capture royalties. Treating all three as identical would obscure the real issue: consent, attribution and manipulation.
The scale of automated supply is already visible. Deezer reported in July that fully AI-generated tracks had exceeded half of its daily new uploads at a June peak, averaging about 90,000 a day. The platform says it labels detected AI albums, excludes fully synthetic tracks from recommendations and editorial playlists, and removes material tied to fraudulent streaming. It also reports that as much as 85 percent of plays associated with fully AI-generated music has been fraudulent.
Industry rules are beginning to catch up. IFPI’s 2026 principles say AI-assisted recordings should qualify for official charts only when the service used is authorized and lawful, the recording is substantially human-made, and its streams are not manipulated. Those tests attempt to protect legitimate experimentation while closing the path from mass generation to artificial popularity.
What visual artists should notice
The music business is an early-warning system for every creative field. The same pipeline can be adapted to images, books, videos and artist identities: generate at scale, attach a recognizable name or style, publish through a low-friction distributor, and use automated engagement to manufacture visibility. Once that material enters recommendation systems, a creator is forced into a costly cycle of detection and takedown.
Artists should make authenticity easier to verify before a dispute begins. Maintain a current list of official channels and authorized distributors. Keep dated masters, high-resolution source files, contracts and release records. Use consistent credits and identifiers. Where appropriate, state whether synthetic voices, avatars or generative tools are authorized. If a suspected impersonation appears, preserve the URL, screenshots, timestamps, account details and audio or visual evidence before filing a report.
Publishers and cultural organizations have a parallel responsibility. They should verify that a work attributed to an artist came through an official source, label meaningful synthetic elements, and avoid amplifying a sensational deepfake while reporting on it. Discovery systems also need provenance signals that distinguish an artist’s own experiment with AI from an unauthorized simulation of that artist.
The important distinction
Sony’s 260,000 takedowns do not prove that generative music itself is illegitimate. They show that scale without provenance can overwhelm a rights-enforcement system designed for isolated infringements. A useful policy response should therefore protect experimentation and access to creative tools while demanding consent for identity imitation, transparent labeling, reliable complaint channels and consequences for fraudulent distribution.
For artists, the immediate lesson is practical: identity is now part of the asset that must be documented and defended. For platforms, removing a fake after it spreads is no longer enough. Prevention, recurrence controls and trustworthy provenance are becoming essential parts of cultural distribution.
ATIM approaches artificial intelligence with curiosity, independence, and balance. We explore its extraordinary potential to empower artists while examining the questions it raises around authorship, copyright, consent, creative identity, transparency, and human expression.
We are neither advocates for AI nor opponents of it. Our focus is how the technology is used—and what that means for artists.
Above all, our commitment is to artists. We look at every development through the lens of what can support, protect, and empower the creative community—identifying opportunities, examining potential risks, and providing artists with the information they need to make informed decisions about their work, their rights, and their future.