Watermarking the Wild West of AI Music

The recent announcement by AI music startup Suno to introduce audio watermarking and fingerprinting technology is a necessary move in response to rampant misuse on its platform. This decision comes as the company faces multiple lawsuits from major record labels, including Universal Music Group and Sony Music Entertainment, which accuse Suno of using copyrighted recordings without permission to train its AI models.

The North Carolina man who pleaded guilty to using hundreds of thousands of AI-generated songs and billions of fake streams to collect over $8 million in royalties is an example of the creative ways scammers have exploited the unregulated space of AI music generation. The lack of oversight has allowed this type of abuse to thrive, with minimal consequences for those involved.

Suno’s new safeguards aim to combat these schemes by introducing audio watermarking and fingerprinting technology that will be durable and resistant to tampering, without affecting the listening experience. However, as Mikey Shulman, co-founder and CEO of Suno, acknowledged in a recent blog post, there is a thin line between aggressively combating fraud and stigmatizing AI-generated music.

The debate surrounding AI-generated music centers on whether it should be clearly labeled or left to the discretion of artists and platforms. Spotify’s decision to partner with Universal Music Group on an AI-generated covers tool suggests that they are leaning towards the latter approach. However, as Suno navigates its legal battles, one thing is clear: the days of easy profits for scammers in the AI music space are numbered.

The GEMA ruling in Germany, which found that Suno violated the copyrights of musicians represented by the country’s music-licensing agency, marks a significant turning point. This decision sends a signal to other companies operating in this space: compliance with copyright laws is not optional. As one expert noted, “the industry has been caught sleeping at the wheel on AI-generated music, and it’s time for companies like Suno to take responsibility.”

The proposed class-action lawsuit over the data breach that exposed the personal information of more than 55 million users also raises questions about Suno’s commitment to transparency. The company allegedly scraped YouTube, Deezer, Genius, and other online sources to train its models, which only adds fuel to the fire.

As the industry evolves, one thing is certain: the need for accountability and clear labeling of AI-generated music will only grow. For now, Suno’s decision to introduce audio watermarking and fingerprinting technology marks a step in the right direction – but it’s just the beginning.

The implications of this development extend far beyond Suno itself, however. As the music industry grapples with the ethics of AI-generated music, questions about ownership, authorship, and compensation will only grow more pressing. Will we see a future where AI-generated music is treated as a distinct category, subject to its own set of rules and regulations? Or will it be assimilated into the existing musical landscape?

For artists, labels, and platforms alike, the stakes are high – but so too are the opportunities. As the dust settles on Suno’s troubles, one thing becomes clear: the Wild West of AI music is giving way to a more regulated era. The question is – what will we make of it?