No AI on the Charts (for Now)

Australia’s record industry will bar wholly AI-generated tracks from its official charts starting next Monday. Under the Australian Recording Industry Association’s new test, a track is eligible only if “humans wrote the song and performed the lead vocal and the primary instruments,” among other requirements. AI-generated music is also ineligible for the ARIA Awards.

The song that forced the rule is a cover of Madonna’s “Like a Prayer” built by an Australian producer with AI-generated vocals and drums. The song has spent 16 weeks in the Australian top 20, peaking at number two in May and sitting at number four on the day the ban was announced. The track has more than 48 million Spotify streams.

The rule is narrower than a full AI ban. Producers can still use AI for mastering, drum machines, and auto-tune, and ARIA says some AI-assisted music remains eligible. The requirement is that humans write the song and perform the lead vocal and the main instruments. There is also a licensing test: eligible music must be made with legal AI services. ARIA chief executive Annabelle Herd said “a chart that rewards unlicensed AI output would undercut the very basis of the recorded music we exist to represent.” That word carries weight; a German court ruled in July that Suno broke copyright law, the first such finding in Europe. ARIA is applying the International Federation of the Phonographic Industry’s July principles, which say a track must be “substantially human made” to qualify for official charts, so every other national chart has the same text ready to adopt.

This is “deja vu all over again.” Back in the 70s, I was one of the first musicians listed in the Local 802 directory as a synthesizer player. The union used to send a representative to my recording sessions to ensure that the producers were not using me (and my truckload of gear) to replace “live” musicians. It was the dawn of the disco era, and we were just starting to use analog synthesizers to play parts that were traditionally played by whole string sections. This was frowned upon (to say the least). Within a few years, analog synthesizer manufacturers started creating dedicated string synthesizers (such as the Cordovox Freeman String Symphonizer and the ARP String Ensemble) for musicians who did not know how to program general purpose music synthesizers. This drove the union crazy. Traditional musicians (especially violinists, violists, cellists, and double bass players) felt these machines threatened their livelihood, which they did in certain pop music genres.

However, synthesizers and sequencers (and ultimately samplers and personal computers) also ushered in a new era of creativity and several new genres of music including Synth-Pop, Hip-Hop, House, Techno, Ambient, Industrial, and EDM. You will not find a modern recording studio that does not use every digital tool at its disposal. If we are being honest, the amount of digital signal processing and commonly used digital production tools makes the idea of “human created tracks” laughable. As a rule, people no longer record performances – they perform recordings. In other words, the soundscapes you are listening to cannot exist in nature; they only exist in the artificial world of electronic music production.

We know how this story will end. Today, those who feel threatened are making their feelings known. In time, no one will care what tools their favorite artists use to create their art.

The anti-AI movement is growing everyday. This new rule is a symptom and, while I know that sometime in the future AI will be accepted as just another technology in our digital lives, we should not (under any circumstances) dismiss the powerful feelings on display. Even if I’m right about the future acceptance of AI, the road to get there is going to be rough.

Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various generative AI models.