AI Music Companies Face Scrutiny Over Unauthorized Training Practices

AI-music companies face renewed scrutiny over unauthorized training practices and allegedly fraudulent streaming, according to a guest post by Jamie Njoku-Goodwin published September 30, 2026.

Njoku-Goodwin, the former CEO of U.K. Music, cited Deezer’s claim that half of new uploads to streaming platforms are now AI-generated and that 85% of streams of fully AI-generated music are fraudulent. The figures were presented as Deezer’s findings, not independently established in the report.

The commentary also points to copyright litigation involving The New York Times, OpenAI and Microsoft. Newly unredacted filings reportedly show Microsoft director of applied science Brent Hecht describing the company’s copying practices as “an astonishing theft of unprecedented proportions.”

Suno has agreed licensing deals with record labels, but Universal Music Group and Sony Music allege that its newest model used outputs from earlier models trained on their recordings. They characterize that process as laundering copyrighted content, a claim whose legal outcome remains unresolved.

Njoku-Goodwin argues that later licensing agreements should not serve as amnesty for unauthorized training. His proposed response is continued copyright enforcement and a rejection of “take without permission first and then settle later” as a sustainable business model.