Artificial intelligence has already changed how music can be created. Now the music industry is confronting a more difficult question: when should AI-generated music be allowed to compete alongside human artists?
Australia has just provided one of the clearest answers so far.
The Australian Recording Industry Association, better known as ARIA, has changed its chart rules so that wholly AI-generated recordings are no longer eligible for the country’s official music charts. Music created by humans with AI used as a supporting tool can still qualify.
The distinction is significant because it moves the debate beyond whether musicians should use artificial intelligence at all. Instead, the industry is beginning to define where the boundary lies between AI as a creative tool and AI as the creator itself.
That boundary could become increasingly important for artists, labels, distributors and streaming platforms.
Australia Draws a Line Around AI Music
ARIA’s updated rules take effect with the chart dated August 31, 2026. Under the new system, a recording developed using generative AI must be “substantially human made” to remain eligible for the ARIA Charts. It must also comply with applicable laws and raise no concerns about streaming or chart manipulation.
ARIA divides the use of artificial intelligence into three broad categories.
AI-generated music is not eligible. This includes recordings where generative AI created the primary creative elements. An AI-generated lead vocal, an important AI-generated instrumental performance or a song generated entirely from a prompt can place a recording in this category.
AI-assisted music remains eligible. Here, humans create and perform the central elements of the recording while generative AI contributes in a secondary role, such as backing vocals or non-primary instrumental elements.
AI used in production is also eligible. ARIA specifically lists applications such as AI mastering, stem separation, reverb and instrument patches among acceptable production uses.
This distinction is crucial. ARIA isn’t saying that using artificial intelligence automatically makes music ineligible. It is asking where the human creative performance sits within the recording.
Read ARIA’s official AI chart rules
The Controversy That Brought the Question Into Focus
The rule change arrives after controversy surrounding Australian DJ Josh Fawaz’s cover of Madonna’s Like a Prayer.
The recording became the most-played song on Australian radio and reached No. 4 on two ARIA charts in July. Producers subsequently raised questions about its use of generative AI and whether that use had initially been adequately disclosed. Fawaz has said he used AI as a tool and had been releasing music before the current generation of AI technology.
Whatever the final interpretation of that particular recording, the controversy exposed a problem that music charts can no longer avoid.
Charts were created to measure the popularity of recorded music. Historically, there was little reason to ask whether the vocalist, instrumental performance or artist identity being measured actually existed in the traditional sense.
Generative AI changes that assumption.
If a song generated largely from a prompt can accumulate millions of streams, should those streams compete directly with recordings built around human performances?
ARIA has now answered that question for its charts: not if the primary creative elements are AI-generated.
This Is Bigger Than One Australian Chart
Australia’s decision matters because it reflects a wider shift taking place across the music industry. Streaming platforms are also becoming more explicit about distinguishing different forms of AI involvement.
Spotify recently announced that AI Persona badges will begin appearing on some artist profiles from mid-September. These badges identify artist identities that may be AI-generated rather than representing real people.
Spotify is making another important distinction, however. The AI Persona badge concerns who the artist claims to be, not necessarily how the music was produced.
For information about the creative process itself, Spotify has introduced separate AI Credits and SongDNA features. Spotify says tens of thousands of AI credits are already being submitted daily.
Even more significantly for music discovery, Spotify says AI Personas will not be included by default in editorial or algorithmic recommendations. A listener could still encounter them after actively demonstrating interest, such as by following the artist.
This suggests the debate is moving beyond simple labelling. Platforms are beginning to consider whether the use of AI should affect visibility, recommendation and competition.
Read Spotify’s announcement about AI Personas
Apple Music Is Moving Toward AI Labelling Too
Apple Music is taking another approach.
According to Music Business Worldwide, Apple has told industry partners that content materially generated using AI platforms such as Suno will receive a “Made With AI” label later this year when the appropriate metadata has been supplied by labels and distributors.
Apple has reportedly operated AI Transparency Tags behind the scenes since March, allowing music companies to identify AI involvement in areas including recordings, compositions, artwork and music videos. The planned change would make some of that information visible to listeners.
There is an important complication. The system relies heavily on labels and distributors accurately declaring AI use.
That introduces another question the industry will need to solve: who is responsible for deciding what counts as AI-generated?
Read the Music Business Worldwide report
The Hard Part Is Defining “Human-Made”
At first glance, separating human music from AI music might sound straightforward. In practice, modern production makes the boundary surprisingly complicated.
Consider a songwriter who writes every lyric and chord but uses AI mastering. Another musician might record all the instruments but generate background vocals. A singer could record a vocal performance that is subsequently transformed using generative technology.
Then there are more extensive uses. An artist might write the lyrics, select the style and repeatedly prompt a generative music system until it produces the desired recording. At the other end of the spectrum, someone could enter a single prompt and publish the first completed song the system produces.
All of these examples involve artificial intelligence, but the degree of human creative contribution is dramatically different.
ARIA’s rules acknowledge this complexity by allowing AI-assisted recordings and production tools while excluding recordings where AI created the primary creative elements.
That may prove more workable than treating “AI music” as one category.
AI-Assisted and AI-Generated Are Becoming Different Categories
This distinction could become one of the most important pieces of terminology in the next phase of the music industry.
AI-assisted describes music where humans remain at the centre of the creative work and AI contributes to part of the process.
AI-generated describes music where generative systems create the entirety or primary portion of the creative elements.
ARIA has formally adopted that distinction for chart eligibility. Spotify, meanwhile, separates the question of AI-generated artist identity from AI involvement in the recording itself.
This is a more nuanced approach than simply attaching an AI label to anything touched by machine learning. Modern music production already contains enormous amounts of technology, so drawing the line at whether a computer was involved would be almost meaningless.
The more relevant question is increasingly becoming: what did the human actually create and perform?
Should AI Music Be Allowed on the Charts?
There are reasonable arguments on both sides.
One argument is that charts should simply measure what people listen to. If millions of listeners voluntarily stream an AI-generated song, excluding it means the chart is no longer a pure representation of listening behaviour. From that perspective, the technology used to create a recording is secondary. Popularity is popularity.
The opposing argument is that official charts are part of a wider professional music ecosystem. Chart positions can generate publicity, awards, career opportunities and commercial value. Allowing unlimited machine-generated recordings to compete with human artists could fundamentally alter what those charts represent.
There is also a question of scale. Generative systems can potentially produce music at a rate no human musician could match. If production becomes effectively unlimited, the economics and incentives surrounding chart competition could change substantially.
ARIA has chosen to preserve human creative performance as part of chart eligibility while still allowing musicians to use AI tools.
That is an important distinction. The policy isn’t human versus technology. It is about determining whether there is still meaningful human authorship and performance at the centre of the recording.
Transparency May Become More Important Than Prohibition
The direction emerging across the industry suggests that outright bans may not become the dominant approach. Instead, we are beginning to see a combination of classification, disclosure and different treatment depending on how AI was used.
ARIA distinguishes AI-generated from AI-assisted recordings. Spotify is separating AI-generated identities from AI involvement in music creation and changing how AI Personas participate in recommendations. Apple Music is preparing visible AI labelling based on metadata supplied through the music distribution chain.
These approaches are different, but they share a common principle: listeners and industry systems increasingly need to know what they are actually dealing with.
That could ultimately be more important than attempting to eliminate AI from music altogether.
What This Means for Independent Artists
For independent musicians and labels, these developments make accurate metadata and transparency increasingly important.
Until recently, declaring the role of AI in a recording might have seemed largely optional. That is changing quickly. AI information can now potentially influence chart eligibility, how a release is labelled and whether an artist identity participates in algorithmic recommendations.
Independent artists should therefore pay close attention to questions asked by distributors about generative AI. Those fields may eventually carry considerably more significance than they do today.
There is also a broader strategic lesson. As generating content becomes easier, identity, creative intent and trust may become more valuable rather than less.
Listeners may increasingly want to know not only what a song sounds like, but who created it and how.
The Music Industry Is Beginning to Choose
The debate around AI and music is often framed as a prediction about the distant future. ARIA’s decision shows that the future has already reached some of the industry’s basic infrastructure.
A recording can now generate legitimate streams and still be considered ineligible for an official national chart because of how its primary creative elements were produced. At the same time, AI-assisted recordings remain eligible.
Spotify is preparing to identify synthetic artist personas and change how they participate in recommendations. Apple Music is preparing more visible AI transparency measures.
The important question is therefore changing. It is no longer simply whether AI can make music. Clearly, it can.
The more consequential question for the industry is now:
When AI makes music, where should that music be allowed to compete, how should it be identified, and what role must humans retain?
Australia has drawn one of the first clear lines. It is unlikely to be the last.
Frequently Asked Questions
Is AI-generated music banned from the Australian charts?
Wholly AI-generated recordings are no longer eligible for the ARIA Charts. However, music that uses generative AI in a supporting role can remain eligible if it meets ARIA’s other requirements.
Can musicians use AI and still enter the ARIA Charts?
Yes. ARIA explicitly permits AI-assisted recordings and the use of AI in production. The recording must remain substantially human-made and comply with the other chart rules.
Does AI mastering make a song AI-generated?
Not under ARIA’s rules. The organization specifically identifies AI mastering as an example of AI used in production, which remains eligible. Stem separation, reverb and certain instrument patches are also permitted.
What is the difference between AI-generated and AI-assisted music?
Under ARIA’s framework, AI-generated music has primary creative elements produced by generative AI. AI-assisted music remains primarily created and performed by humans, with generative AI used for secondary elements.
Is Spotify banning AI music?
No. Spotify’s AI Persona badge concerns AI-generated artist identities, not all music created using AI. Spotify says AI Personas will be excluded by default from editorial and algorithmic recommendations, although listeners can still actively choose to engage with them.
Will Apple Music label AI-generated songs?
Apple has reportedly told industry partners that materially AI-generated content will receive a “Made With AI” label later in 2026 when the appropriate AI metadata is supplied by labels and distributors.
Sources and Further Reading
ARIA. ARIA Charts Set Eligibility Rules for Recordings Made With AI. August 2026.
Read ARIA’s official announcement
Spotify. Introducing a New Label for AI-Generated Artist Identities on Spotify. August 2026.
Read Spotify’s announcement
Reuters. AI-generated music barred from Australian charts after Madonna cover controversy. August 2026.
Read the Reuters report
Music Business Worldwide. Apple Music prepares AI labelling for music. August 2026.
Read the Music Business Worldwide report