Last month, acclaimed jazz pianist and composer Jason Moran received an unusual call from his friend, bassist Burniss Earl Travis. Travis had spotted what appeared to be Moran’s new album on Spotify and questioned its authenticity. Moran, who avoids the platform and distributes his work solely through Bandcamp, investigated and discovered a profile under his name featuring releases from his past label, Blue Note Records, which holds rights to his earlier recordings. Among them was a recent EP called For You, with cover art showing a young woman in a rainy, anime-inspired scene. Upon listening, Moran noted the absence of piano and described the tracks as indie pop, far removed from his style. He took steps to have the unauthorized content removed. Moran joins a rising number of artists affected by apparent AI-generated imitations on streaming services. This issue has impacted over a dozen prominent figures in jazz, indie rock, and even rapper Drake. Artists find the influx of such artificial material aggravating and eerie, with Moran comparing it to a Black Mirror episode involving a duplicated celebrity persona disrupting the real person’s life. Spotify recognized the challenge last September, stating it had eliminated over 75 million spam tracks in the prior year and was enhancing artist safeguards, including stricter impersonation guidelines. In a recent update, the company announced a forthcoming feature allowing artists to review and approve or reject uploads associated with their names, emphasizing identity protection as a key focus. A Spotify representative explained that the platform uses various measures like automated detection, manual oversight, and reporting mechanisms to block improper content, claiming uniqueness in offering this artist control tool. However, Moran, previously the jazz artistic director at the Kennedy Center, believes these efforts fall short, particularly since AI material often evades internal checks and the problem persists. He worries about the extra burden on non-Spotify users like himself and on deceased artists. He questioned how figures like John Coltrane or Billie Holiday could contest fabricated releases, such as a supposed rediscovered 1952 Paris concert, noting their inability to intervene. The spokesperson clarified that representatives of deceased artists’ estates can access the new tool with an account, while for others without profiles, Spotify depends on its built-in monitoring. After Travis alerted him, Moran shared a video on social media about the incident, prompting responses from numerous artists reporting similar experiences, some ongoing for years. In jazz, affected individuals include pianist Benny Green, saxophonist Antonio Hart, drummer Nate Smith, band Hiatus Kaiyote, and vocalists Dee Dee Bridgewater, Jazzmeia Horn, and Freddy Cole, sibling of Nat King Cole. Moran highlighted the risk, suggesting a fake album under a name like Frank Ocean would attract streams regardless of origin. Last October, reports emerged of indie rockers Luke Temple and Uncle Tupelo, plus late artists Sophie and Blaze Foley, having their profiles compromised by AI. In December, after Australian group King Gizzard and the Lizard Wizard pulled their catalog from Spotify, an AI copycat named King Lizard Wizard emerged with matching track names and subpar generated visuals. Morgan Hayduk, co-CEO of fraud detection firm Beatdapp focused on music streaming, indicated that AI acts as a catalyst for such issues.
