Suno Sparks Music Copyright Nightmare, Prompting Industry Alarm

Sun o’s “Premier” plan, which costs $24 a month, includes access to Sun o Studio—a tool that lets users upload an existing track and ask the system to generate a new version. According to Terrence O’Brien of The Verge, the platform’s copyright filters are supposed to block any use of protected material, but they can be bypassed with simple audio‑processing tricks such as halving or doubling the playback speed in a free editor like Audacity, then adding a burst of white noise at the start and end of the file. When the altered file is re‑uploaded to Sun o Studio, the system fails to recognise the source, restores the original tempo, strips the noise, and proceeds to produce an AI‑generated “cover” that is “alarmingly close” to the original recording (The Verge).

The practical upshot is that a user can take a hit song—examples cited include Beyoncé’s “Freedom,” Black Sabbath’s “Paranoid,” and Aqua’s “Barbie Girl”—and, after a few seconds of manipulation, generate a new track that retains the original’s instrumental arrangement and vocal timbre. O’Brien notes that the AI’s output varies with the model version: models 4.5 and 4.5+ make only minimal changes to the sound palette, while model v5 adds more aggressive stylistic alterations, such as “chugging guitar and galloping piano” to Beyoncé’s track or turning a Dead Kennedys song into a “fiddle‑driven jig.” In either case, the result can be exported and uploaded to streaming services, where it may be mistaken for an alternate take or a previously unreleased B‑side by casual listeners (The Verge).

The loophole extends to lyrics as well. Sun o’s policy blocks direct copying of copyrighted text, but O’Brien discovered that “extremely minor changes” to the spelling of words—e.g., swapping “rain on this bitter love” for “reign on” or “tell the sweet I’m new” for “tell the suite”—allow the AI to generate vocal tracks that closely mimic the original performance while evading the filter. When users paste full lyrics from a site like Genius, the system flags the content and produces “gibberish vocals,” yet the same minimal edits slip through undetected (The Verge).

Industry observers are alarmed because the technology makes it trivial to flood digital platforms with near‑identical replicas of high‑profile songs, potentially siphoning royalties and complicating rights enforcement. The Verge’s report frames the situation as a “music copyright nightmare,” noting that the ease of manipulation—requiring only free software and a modest subscription—could lead to large‑scale abuse. While Sun o declined to comment, the article underscores that the company’s own safeguards are “incredibly easy to fool,” raising questions about the adequacy of current AI‑generated content filters and the liability of platforms that host such material.

For record labels and rights holders, the emergence of this loophole forces a reassessment of monitoring strategies. Traditional fingerprinting services may struggle to differentiate AI‑generated imitations from authentic recordings when the underlying arrangement and vocal characteristics remain largely intact. The potential for “monetising these uncanny valley covers” by uploading them to streaming services could erode revenue streams and create legal exposure for both the AI provider and the distributors that inadvertently host infringing content. As O’Brien’s findings illustrate, the problem is not merely theoretical; it is reproducible with “minimal effort” and “some free software,” suggesting that the threat could scale quickly if left unchecked.

Analysts note that the episode reflects a broader tension in the AI music market: the promise of rapid, low‑cost content creation versus the risk of widespread copyright infringement. Sun o’s case may prompt regulators and industry groups to push for stricter verification protocols, while platforms that host user‑generated music may need to invest in more sophisticated detection tools. Until such measures are implemented, the “music copyright nightmare” described by The Verge is likely to persist, challenging the industry’s ability to protect intellectual property in an era of increasingly accessible generative audio technology.