Audio AI expert Ircam Amplify unveils groundbreaking software to identify AI-generated tracks

02 May 2024 - Press release

First of its kind independent tool brings greater transparency on AI music to the global music industry

 

Ircam Amplify, the global leader in technology solutions that analyse, process and edit audio, today launches its innovative AI-Generated Detector tool to empower record labels, publishers, performing rights and collective management organisations, distributors, streaming platforms and download stores to accurately identify AI-generated music.

The technology behind the first of its kind AI-Generated Detector comes to market as hundreds of thousands of new music files created by AI-powered tools are uploaded daily on music streaming platforms, such as Apple Music, Spotify, and Amazon Music. In 2023 alone, an estimated 170 million tracks were created using AI, posing a risk to human creativity, infringing copyright and diluting compensation for music creators.

Ircam Amplify - a branch of IRCAM-Centre Pompidou, the largest audio lab in France and a major centre of music creation - has developed a new AI-tagging system embedded in the AI-Generated Detector to allow stakeholders to easily identify “artificial” music files in a fully automated manner. The cutting edge technology can screen up to 5,000 tracks in less than a minute with an impressive accuracy rate of 98.5% in identification. 

Nathalie Birocheau, CEO of Ircam Amplify, said: “Generative AI is reshaping the creative landscape, especially in the music sector, where everyone can create music using AI tools. This is, quite rightfully, a concern for all those who are involved in the creation, distribution and revenue-sharing of music.

“Our AI-Generated Detector is a game changer. It will provide transparency where there was opacity. It will give a strategic advantage to the users of the detector and help them make informed decisions about music content. This is a much needed development for the market and we are thrilled to be at the forefront of this new technological development.”

Romain Simiand, Chief Product Officer at Ircam Amplify, added: “Leveraging our Music Information Retrieval (MIR) and audio AI expertise, we trained a dedicated algorithm through a deep neural network on both humanly created and publicly available AI-generated music datasets. The detection training proved so specific to each model that we were able to state which one has been used to generate the tracks.”

The AI-Generated Detector follows the successful launch last year of Ircam Amplify's flagship product, Stereo to Spatial, which offers a seamless and affordable spatialisation process for record labels and digital distributors.

In light of music companies and streaming platforms strive for quality, Ircam Amplify is providing a comprehensive suite of tools designed to ensure that high standard requirements are met at every step of the supply chain, with Quality Check and Spatial Control complementing the AI-Generated Detector. Other innovative analysis and editing AI products include Music Tagger, Vocal Separator, Loop Extractor, Infringement Detector, Cover Detector and Speech to Text Generator.  

Demos of the AI-Generated Detector will be available during the Music Biz Conference, held in Nashville from 13-16 May 2024, where delegates will be able to see the product in operation for the first time.