Who really 'owns' AI generated work and who gets the money? It’s a question that’s been debated ever since this particular genie escaped its bottle. So you installed the software, you typed the prompt and whatever comes out of the other end is yours, right? Wrong?
In a potential landmark decision for AI – and its place in generative, artistic creation going forwards – the US copyright office just decided that work produced by AI can’t be copyrighted. That is, it’s impossible to own something that AI created no matter if it was you that pushed the buttons and conjured it into being.
AI-based tools however – such as those used to reconstruct lost performances or to enhance or modify created works to achieve the ‘impossible’ – are still permitted and work containing them can still be attributed to a creator and ‘owned’.
So what does it all mean?
Well, the new ruling is clearly bad news for anyone currently (or planning to) get rich while AI does all the hard work. Most specifically the platforms and services that were hoping to make money from training AI on existing work and passing it onto users to allow them to run riot.
Meanwhile, the subtle distinction regarding ‘AI tools’ is good news for those offering new technology, allowing progress in the fields of sound enhancement and synthesis to take place.
But, most importantly it’s good news for musicians and artists creating original human-derived works.
The fear had been that with AI’s inevitable rise, countless millions of new songs, artwork, video, photography and more would be dumped onto the market, effectively devaluing and swamping human-created work, making their ‘real’ creations indistinguishable in a sea of clones and role of the genuine creator, ultimately, unnecessary.
This new obstacle in the way of that particular nightmare means that – while not eliminating the availability of such technology or in any way preventing anyone from using it to create – to do so would ultimately be financially pointless.
Plus, AI gains its intelligence through being trained on existing human-derived knowledge and experience. And while AI swallowing up physics textbooks like there’s no tomorrow has rattled a few cages, AI consuming recorded work by established artists (and potentially subsequently producing soundalikes that are as good or better than the real thing) is clearly something that all of today’s biggest stars would like to put the brakes on – or at least get paid for.
Recently such high profile acts as Paul McCartney and Elton John have spoken out about AI and its potential impact on the revenues available to artists, and subsequently, with no livelihood to be made and no chance of recognition in a sea of AI-produced clones, the number of artists that could choose to exist in the future.
Imagine no more music. It’s easy if you try
The new report finds that: “Based on the fundamental principles of copyright, the current state of fast-evolving technology, and the information received… the Copyright Office concludes that existing legal doctrines are adequate and appropriate to resolve questions of copyrightability.
“Copyright law has long adapted to new technology and can enable case-by-case determinations as to whether AI-generated outputs reflect sufficient human contribution to warrant copyright protection.
“In many circumstances these outputs will be copyrightable in whole or in part – where AI is used as a tool, and where a human has been able to determine the expressive elements they contain. Prompts alone, however, at this stage are unlikely to satisfy those requirements. The Office continues to monitor technological and legal developments to evaluate any need for a different approach.”
All in all it’s a bit of a breakthrough and a bit of a kick to the teeth for platforms such as Suno, the AI-powered song generation platform endorsed by producer Timbaland which aimed at mainstream music creation and wrench it out of the hands of creators.
The message is, sure you can make a song with AI prompts – knock yourself out – but they’ll only ever go on a pile of abandoned, purposeless, valueless soundalikes.
While, in a victory for common sense (in a music business driven by innovations such as the electric guitar, synths, samplers, drum machines, DAWs and ever more impressive sound libraries and plug-ins) AI-based tools for making music are still permitted.
It looks like a win win… But when does an 'AI tool' cross the line into being an 'AI fake'? Let’s see how long it takes before someone bends the rules…
