Warner Music signs distribution deal with AI generated music app Endel

AI and copyright is one of the hottest topics of the time in IP at the moment. This is hardly surprising since it raises so many interesting questions about creativity and ownership. As Prof. Bernt Hugenholtz argued at the March Alicante Congress on AI and IP [reported here] that copyright protection – if any - for AI-generated works needs to be balanced against the overarching goals that such protection aims to achieve.

However, this has not stopped Warner Music signing a bundle of code to create 20 new albums this year! Warner Music is an American multinational entertainment and record label, the third largest in the global music industry, with artists from Ed Sheeran, Coldplay, Madonna to Led Zeppelin. The latest to be "signed" by the multi-billion dollar company, is an algorithm. 

Endel, developed by a start-up based in Berlin, creates tailor-made custom sound frequencies based on personal user inputs such as weather, time of day, location, and biometric details such as heart rate. 

 

Cat Sampling
Image: Andy Miccone

Evidently, Endel is not signatory to the contract with Warner, as such. The company has agreed a 50/50 distribution deal, covering a total of 20 albums that will be released throughout 2019 with Warner Music. 

This is not the first time AI-generated music has created distribution deals. Aiva Music is a composition algorithm that famously became the first AI to register with a collecting society (SACEM) and recently partnered with Believe Distribution (owned by Song Records) to release its latest album. Sony also has its Flow Machines project, which involves a algorithmic composition tool that is formally credited as a songwriter, producer, instrumentalist and/or vocalist in all of the tracks’ liner notes for its debut album 'Hello World'. However, the creators also include a list of human contributors who provided songwriting, instrumentation, mixing and mastering support.

The Endel-Warner deal is a step forward in that there are no human collaborators in the generation of the new sounds. Nevertheless, a human - aside from obviously creating the AI - also had to, input sounds and data into Endel. Interestingly Stavisky describes the work as being "generated based on different combinations of inputs" rather than created. 

Some of these inputs, or instrumental stems, were created by Endel's co-founder and sound designer Dmitry Evgrafov. Each sound is then allocated metadata according to certain parameters which the app can read and use to generate a soundscape. So, whilst it might seem that the sounds are created with a click of a button, Stavisky explained that it took "1.5 years of work developing our algorithm and creating and tagging the stems.” 

In terms of copyright ownership, on a theoretical level some argue that the creator of the AI might be the owners of the outputs, others suggest that it could be the AI system itself. Other potential owners could be the creator of the "inputs" on the basis that this is the personality being expressed. Or, perhaps, the investor of the AI project on an economic justification of remuneration as encouragement.

In any event, for Warner and Endel, it was practical issue. Stavitsky said that when Warner asked for the songwriter information in order to register the copyright of the music, they decided to list the co-founders and software engineers, saying “I am now credited as a songwriter even though I have no idea how to write a song."

 

Endel: "Personalized sounds to help you focus and relax"

So, whilst we are still discussing what we think the outcome should be in theory, in practice some are marching ahead on the basis that the copyright holders are the company founders and the AI engineers. At this stage, given the extensive skill, labour and effort that went into the development of Endel it might not be so controversial. But what happens if [when] it is a self-learning machine that doesn't require as much human effort?

So many other questions come to mind - is Endel sampling? Do they need a licence? Who would be liable if Endel created an infringing piece of music? All of the registered copyright holders? The specific engineer who input a copyright protected work? But it is only if Endel uses a substantial part of that work in a new song that it would be infringing. Are the engineers able to programme the system not to take a substantial part? As we well know, its not about quantity in which case that might be viable, but since it is something decided on "quality", on a case by case basis, it's not so straight forward.

The future for copyright and AI remains to be seen, but it appears that the time is ripe to be discussing such issues!