What the Algorithm Thinks Australian Music Is
On US taste reproduced as global norm, the specific cost to Australian artists, and why local curation isn't nostalgia
Research published in 2025 examined the composition of AI-generated playlists across multiple markets and found that they heavily relied on global listening patterns, meaning primarily US listening patterns, because the US generates a music footprint large enough to function as a proxy for what the algorithm understands music to be. AI-generated playlists in Australia looked very similar to AI-generated playlists in every other country studied. They looked like the US. Frankly… I’m bloody tired of being compared with the US… especially right now.
'AI recommendations accentuate US dominance by reproducing US tastes as global norms,' researchers found. 'Our study showed the composition of AI playlists in all countries is very similar to those of the US.'(1) The AI playlists were also less likely than editorial playlists to surface diverse or regionally specific music. Which means that when an Australian listener opens Spotify and lets the algorithm find them something new, what they are most likely to find is music that the American market has already decided is popular. The algorithm is not neutral. It has a taste. The taste is American.
This is not a new problem. The domination of English-language American music in global streaming has been a consistent feature of the platform era, and it predates AI. But AI-generated recommendations have the specific property of systematically reproducing and amplifying the existing bias rather than occasionally departing from it the way a human editor with local knowledge might. An algorithm trained on global streaming data learns that what the world listens to is what the largest market listens to. An Australian editorial curator knows that what Australians listen to, and what Australians make, is not the same as what Americans listen to. The algorithm does not know this. It does not have a concept of Australia except as a node in a global stream count.
For Australian independent musicians, this has a measurable consequence. Discoverability on major streaming platforms operates primarily through algorithmic recommendation. Editors at major labels can negotiate playlist placement for their artists through commercial relationships with the platforms. Independent Australian artists do not have those relationships. Their path to new listeners runs through the algorithm. The algorithm has been documented to underperform for regionally specific and culturally specific music.
The music is there. There is no shortage of original, inventive, culturally specific Australian music being made. Perth alone has produced musicians of international quality who have largely built their careers in other cities because the infrastructure to sustain them here was insufficient. Western Australia loses 28% of its dedicated live music venues between 2010 and 2024. The state spends $0.72 per person per year on contemporary music, the lowest per capita investment in the country. The median annual income for independent musicians is $14,700. The algorithm reflects the conditions it finds. It does not create the conditions. It certainly will not fix them.
The argument for local curation — human beings with local knowledge making recommendations informed by where music was made, who made it, and what community it speaks to — is sometimes dismissed as parochialism. The preference for locally specific music over globally popular music treated as a sentimental attachment to the familiar rather than a principled argument about what streaming infrastructure should do.
Recommending Australian music to Australian listeners is not parochialism. It is accuracy. If I am in Perth and I want to find music that emerged from the same physical and cultural landscape I inhabit, music made by people who understand what a Perth summer evening feels like, music that references the specific geography and social texture of this place, the algorithm trained on global streaming data is a less useful tool for that purpose than a human editor who lives here. Obviously. The algorithmic recommendation is not more neutral than local curation. It is differently biased — toward scale, toward English-language American popular music, toward whatever the largest market has already decided is popular — and those biases are not obvious to the listener because the algorithm presents itself as a personalised recommendation rather than a geographically skewed one.
The Pack's geo-fenced local discovery feature is being designed to address exactly this gap. The technology identifies music made locally and streams it to local listeners and businesses, creating the conditions for discovery that the major platform algorithms are documented to suppress.(2) This is not a boutique feature for listeners who already know they want local music. It is a correction for a systematic bias in the recommendation infrastructure, made by a platform whose governance structure is answerable to the musicians most affected by that bias.
The research finding that AI playlists reproduce US taste as global norm is an empirical description of an existing problem. It is also a description of what human curation with local knowledge is designed to correct. Whether an Australian listener discovers Australian music is not determined by the quality of the music. There is excellent Australian music being made at every level of the industry. The question is whether the infrastructure is designed to surface it.
At present, largely, it is not. The algorithm thinks Australian music is Drake, Beyonce, Chappell Roan, Taylor Swift. We know otherwise.
Notes
Peer-reviewed research on AI playlist algorithmic bias and US dominance in Australian market. https://www.apraamcos.com.au/about-us/news-and-events/ai-in-music-report
The Pack Music Co-operative — platform overview and geo-fencing technology. https://www.thepackmusic.com.au
If you've made it this far, you probably care about where music is headed.
So do we — that's why we built something different. The Pack Music Co-operative is Australia's first musician-owned streaming platform: cooperative-governed, human-curated, and built on the radical premise that the people who make the music should own the infrastructure that distributes it.
Join the Pack — become an early adopter member, support our crowdfunding campaign, or lend your voice as an Ambassador: 👉 packmusic.au/join-the-pack
Back the campaign — every dollar goes directly toward getting us to launch:
Read our story — where we came from, why we built it, and what we believe: 👉 packmusic.au/who-we-are
Say hello — we genuinely want to hear from you: 👉 packmusic.au/contact