Loophole Lets AI Songs Infiltrate Official Spotify Page

On September 10, 2026, a new song titled “Riding High” was released on the verified Spotify page of Lathe of Heaven, a punk band based in Brooklyn, New York. The song differed in style from the band’s past work, the vocals were not those of the lead singer, and the artwork was unrelated to their traditional emblem. According to reporting by Emanuel Maiberg of 404 Media, the journalist disclosed that he had produced the song by giving simple instructions to the music-generation AI Udio. Via a music distribution service, the journalist linked the song to the official artist name without the band’s permission. The song reportedly appeared under the same band name not only on Spotify but also on Apple Music, Tidal, and Amazon Music.

This demonstration is a case that reveals a structural loophole in music distribution. While the loophole itself has long existed, it is now combined with the mass generation of low-quality music by generative AI, and abuse is said to be spreading. Third parties exploit the reputation of real performers to earn royalties based on streams. Listeners end up encountering unrelated generated material instead of recordings by the performers they expected. The harm goes beyond mere misregistration; it undermines trust in the artist name and distorts royalty distribution.

Demonstration of Releasing an AI Song Under a

Real Band’s Name

Journalist Maiberg generated the song with only simple instructions to Udio. There was no songwriting, performance, or recording process, and the artwork did not follow the previous style. Even so, it was listed as a new release on the official artist page. While dedicated listeners would have noticed something was off, the system did not detect the anomaly. This is because registration data submitted to the distributor was directly merged into the artist name on each platform.

According to the report, the journalist aimed less at targeting a specific band than at reproducing the loophole. He explained that after the release, he notified the relevant parties and took steps to have it removed. If a malicious actor used the same method, they could accumulate streams and earn royalties before removal. The fact that reproducing the technique required no advanced skills is significant. The spread of generative AI has reduced the cost of forging recordings to almost zero. While the expanding application of generative AI can also be seen in moves such as Microsoft Announces Solara with AI Agent-Native OS, in the music field it has appeared in the form of unverified infiltration.

This demonstration calls into question the meaning of the verified badge. Verification has been understood as a mark of identity confirmation. In this case, however, the verified page itself functioned as a container for unrelated audio. Users cannot judge the provenance of a recording based on the badge. The platforms’ display design and review systems need to be reconsidered.

Exploiting How Music Distributors Work

Music distributors are intermediaries between artists and platforms. Artists do not register with each platform individually, but register once with a distributor. The distributor delivers recordings to multiple platforms such as Spotify and Apple Music. It also handles catalog management and royalty aggregation. It is an essential infrastructure for independent artists.

The problem lies in lax verification of artist names. Even if someone registers under the same or a similar name as an existing artist, the distributor may not strictly verify their identity. Platforms trust the supply from distributors and link the material to the existing artist page. As a result, third-party recordings are mixed into the official catalog. The technique shown in the report exploits this weakness in the matching process.

This structure has long been known. Abuse was also possible through manual imitation or unauthorized re-recordings. Generative AI has made it easy to mass-produce recordings, changing the economics of abuse. A large number of songs can be generated in a short time and submitted under many names. By accumulating streams before detection, even small royalty payments can be collected. The gap is widening between distributors’ review costs and perpetrators’ generation costs.

profiles that aren’t actively managed can be especially vulnerable

The above is the gist of Spotify’s response to the journalist. Artist names that are not actively managed are said to be especially vulnerable. The fact that whether harm occurs depends on whether management is in place shows a flaw in the infrastructure. A design that relies on monitoring by artists is not sustainable.

Why Deceased and Small Artists Are Targeted

This technique tends to concentrate on unknown artists and deceased persons. Well-known artists have robust management systems and detect anomalies early. Small artists cannot spare staff to monitor their catalogs. For deceased persons, confirmation by the individual is impossible. Neglected catalogs become convenient targets for perpetrators.

Journalist Maiberg last year reported the case of Blaze Foley, a local American musician. Foley died in 1989, but new songs appeared under his name on Spotify. The songs were reportedly believed to be AI-generated. In the latest reporting, he confirmed similar phenomena in dozens of names attributed to deceased persons. In the catalogs of some jazz musicians, generated images unrelated to the deceased were used as artwork. In one case, a generated image of a fictional white person was reportedly used as the artwork for a deceased Black person.

Deceased artists’ catalogs are also cultural records. If unrelated generated material is mixed in, it contaminates provenance. Listeners cannot distinguish legitimate reissues from disguised generated material. Research and investigation costs for researchers and reissue labels also increase. Trust in small catalogs is directly linked to trust in platforms as a whole.

The Accountability Gap Between Platforms and

Distributors

Platforms and distributors shift responsibility onto each other. Platforms say that screening suppliers is the distributors’ role. Distributors say that final publication decisions rest with platforms. According to the report, neither side has presented a fundamental solution. While the existence of the loophole is shared, it remains unclear who should fix it.

Some see technical countermeasures as not difficult. For example, stronger verification of artist names at the time of new registration has been suggested. Additional confirmation when a name matches an existing one would be effective. Audio fingerprinting and artwork similarity detection could also be used. Holding royalty payments and establishing appeal channels are also essential. All of these require cost and effort, but implementation is possible.

At present, the cost burden falls on artists and listeners. Artists spend time on monitoring and removal requests. Listeners lose time and attention to incorrect recordings. Royalties may flow to perpetrators. This is a structure in which the businesses operating the infrastructure do not bear the costs. It is a stage where the institutional design needs to be reviewed.

Editorial Opinion

On short-term impact. This demonstration is expected to prompt distributors to strengthen identity verification over the next three to six months. Platforms are also expected to tighten the conditions for merging into existing artist names. Meanwhile, perpetrators may shift their targets to deceased persons and neglected catalogs. While this cat-and-mouse game continues, the burden of removal requests is expected to remain on small operators.

On the long-term perspective. In one to three years, provenance guarantees for recordings are expected to become a requirement for distribution infrastructure. Linking audio fingerprints to registrant information is expected to become standardized. Catalogs without guaranteed provenance are expected to decline in value for reuse and as training data. Preserving music catalogs as cultural records is expected to become a common challenge for the industry.

Questions from the editors. What the verified badge guarantees is now in question. Who should bear the cost of screening suppliers is also at issue. The question remains whether an infrastructure that allows generated material to infiltrate can be called a platform. Attention is focused on whether the industry can show procedures to restore user trust.

References

• “I Hijacked a Real Artist’s Spotify with AI Music. It Was Disturbingly Easy”, by Emanuel Maiberg — 404 Media, 2026-09-17T13:33:52.000Z (ARR)
• Source URL: https://www.404media.co/spotify-ai-music-scam/