Why it is so important to identify them (guest column)

A few weeks ago, a member of my team at Ircam Amplify logged into one of the many AI music generators on the internet and entered a short prompt for a song. Within minutes, a new track was generated and promptly uploaded to a distribution platform. In just a few hours, this song, which involved no human creativity, was available on various streaming platforms. We worked diligently to remove the track from everyone, but the experiment made an important point.

It’s that simple now! I don’t want to make a judgement here about whether AI-generated music is good or bad – we are neutral from that perspective – but we think it’s important to stress that while the process is simple and inexpensive, there are currently no safeguards in place to ensure that consumers know if the music they are listening to is AI-generated. As a result, they cannot make an informed decision about whether to listen to that music.

As digital platforms are flooded with AI-generated songs, streaming services require enormous technological resources to handle the volume of tracks, diverting attention from promoting music created by “human” artists and diluting the royalty pool.

Like it or not, AI is here to stay and more and more songs will find their way onto streaming platforms because the process is so quick and easy. We already know that there are AI-generated music farms flooding streaming platforms. Deezer recently removed over 25 million tracks and it’s reasonable to assume that a significant portion of those were AI-generated.

In the interest of transparency, consumers certainly have a right to know whether the music they listen to is a genuine product of human creativity or derived from computer algorithms. But how can AI-generated tracks be easily distinguished? Solutions already exist. At Ircam Amplify, we offer a range of audio tools, from spatial sound to vocal separation, covering the entire audio supply chain. One of the latest technologies we have launched is an AI-generated detector designed to help rights holders and platforms identify AI-generated tracks. Through a series of benchmarks, we were able to determine the “fingerprints” of AI models and apply them to their output to identify tracks that come from AI music factories.

The purpose of any solution should be to support the entire music ecosystem by providing a technical answer to a real problem while contributing to a more fluid and transparent digital music market.

Discussions around transparency and AI are gaining momentum around the world. From Tokyo to Washington, DC, Brussels to London, policymakers are considering new laws that would require platforms to identify AI-generated content. This is the second recommendation of the report “Artificial Intelligence and the Music Industry – Master or Servant?” recently published by the UK Parliament.

Consumers are demanding it too. A recent UK Music survey of more than 2,000 people, commissioned by Whitestone Insight, strongly found that more than four in five respondents (83%) believe that when a song is created using AI technology, this should be clearly labelled.

A survey conducted by Goldmedia on behalf of the collecting societies GEMA and SACEM in 2023 also found that 89 percent of the collecting societies’ members expressed a desire for clear labeling of AI-generated music tracks and other works.

These staggering numbers show us that concerns about AI are widespread in creative circles and shared by consumers. There are numerous calls for the ethical use of AI, mostly coming from rights holders – artists, record labels, music publishers, collecting societies, etc. – and transparency is usually at the heart of these initiatives.

Simply put, if there is AI in the recipe, it should be labeled. If we can work together to find a way to ensure that AI-generated music is identified, we will have made great progress toward transparency.

Currently CEO of Ircam Amplify, Nathalie Birocheau is also a certified engineer (Centrale-Supélec) and former strategy consultant who has led several major cultural and media projects, notably at the Maison de la Radio. In 2016, she became deputy director of France Info, where she led the creation of the global media Franceinfo.