Artificial intelligence is rapidly changing how music is created, released and discovered. Now, one of the world’s major national chart systems is drawing a clearer line between AI-assisted music and music generated primarily by AI.
The Australian Recording Industry Association (ARIA) has updated its Charts Code of Practice to make wholly AI-generated recordings ineligible for the ARIA Charts. The new rules take effect with the ARIA Chart dated 31 August 2026, published on 28 August. Importantly, ARIA is not banning artists from using AI as a production or creative tool: music that remains substantially human-made can still qualify.
For musicians, producers and labels, this is an important development because it shows how the music industry is beginning to establish a formal distinction between human-led music made with AI assistance and music where AI is responsible for the primary creative contribution.
What Exactly Has ARIA Changed?
ARIA has introduced specific eligibility rules for recordings developed using generative AI.
Under the updated rules, an AI-involved recording can be eligible for the charts when it:
1. Is substantially human-made
2. Does not raise concerns about stream or chart manipulation
3. Complies with applicable laws, including copyright and related-rights requirements.
This means the headline “ARIA bans AI music” needs some clarification.
ARIA has not banned the use of AI in music.
Instead, it has made wholly AI-generated recordings ineligible for its official charts, while allowing AI to remain part of a human-led production process.
ARIA’s own FAQ makes the distinction particularly clear: a recording created by humans that uses generative AI can remain eligible, while a recording where generative AI creates the recording or the primary part of its creative elements is not eligible.
AI-Generated vs AI-Assisted: What Is the Difference?
One of the most important parts of ARIA’s new policy is the distinction between AI-generated and AI-assisted music.
1. AI-Generated – Not Eligible
ARIA considers a recording AI-generated when generative AI produces the entirety, or the primary portion, of its creative elements.
Examples include:
* An AI-generated lead vocal
* An AI-generated key instrumental performance
* A track generated entirely from a text prompt
* Music where AI creates the primary musical elements
ARIA specifically says that an AI-generated lead vocal makes a recording ineligible, just as an AI-generated key instrumental component would.
So, for example, if a musician writes a prompt and an AI system generates the complete song—including the main vocal and primary instrumental performance—the resulting recording would not qualify for the ARIA Charts.
2. AI-Assisted – Eligible
AI assistance is treated differently.
A recording can remain eligible when humans are responsible for the main creative work and AI is used for supporting or expressive elements.
For example:
Human lead vocal + AI backing vocals → Eligible
ARIA explicitly identifies a human lead vocal supported by AI-generated backing vocals as AI-assisted, rather than fully AI-generated.
The same principle can apply when humans write and perform the main song while generative AI contributes smaller elements.
The important factor is that human creativity and performance remain at the centre of the recording.
What About AI Mastering, Stem Separation and AI Instruments?
This is where the policy becomes particularly relevant to modern producers.
ARIA has clarified that several AI-enabled production techniques remain eligible, including:
* AI mastering
* AI drum machines
* Stem separation
* AI-assisted reverb
* AI instrument patches that are subsequently played by a person
These are considered forms of AI in production, rather than AI-generated music, when the underlying creative performance remains human.
This distinction matters because AI is increasingly embedded inside normal music-production software.
Using an AI-powered mastering tool doesn’t suddenly make a human-produced song an “AI song.”
Similarly, using stem separation to isolate vocals or instruments doesn’t mean the resulting recording has lost its human authorship.
The policy is essentially asking a bigger question:
Who is actually responsible for the creative performance at the centre of the recording?
Why Is ARIA Doing This Now?
The timing is closely connected to the extraordinary growth of fully AI-generated music.
Deezer reported in July 2026 that it was receiving nearly 90,000 fully AI-generated tracks every day, with AI-generated music accounting for more than 50% of all new daily uploads at peak levels in June 2026.
But the huge volume of AI uploads does not translate into an equivalent share of listening.
Deezer said fully AI-generated music represented only around 1–3% of total streams, despite its enormous upload volume. Deezer also reported that it detected more than 13.4 million AI-generated tracks during 2025.
The platform has also linked AI-generated music with streaming-fraud concerns. Deezer reported that up to 85% of streams on fully AI-generated tracks were identified as fraudulent in 2025, compared with 8% across its entire catalogue.
This is one reason the conversation has moved beyond simply asking:
“Is AI music good or bad?”
The industry is increasingly asking:
Who made it? Was the AI use authorised? Is there meaningful human creativity? And are the streams genuine?
ARIA’s Rules Are Part of a Global Shift
ARIA’s decision is not happening in isolation.
On 30 July 2026, the International Federation of the Phonographic Industry (IFPI) announced global principles for determining whether recordings developed using generative AI should be eligible for official music charts.
IFPI’s framework establishes a broader set of principles.
Under the framework, recordings developed using generative AI should meet requirements including:
* The AI service should be properly authorised and lawful.
* The recording should be substantially human-made.
* There should be no stream or chart manipulation concerns.
* Applicable copyright, related-rights and personality-rights laws should be respected.
* The recording should comply with the AI service’s terms.
* AI use should be appropriately signalled to consumers where required.
IFPI said it would apply the principles across the charts it directly manages and work with its national network to extend the framework to more than 20 additional official chart programmes.
ARIA is one of the national chart organisations implementing this broader direction.
ARIA Has Also Adopted AI Labelling Standards
The chart rules are part of an even broader industry movement toward transparency.
On 10 July 2026, IFPI, RIAA, A2IM, WIN, IMPALA, The Grammys, SAG-AFTRA and the Human Artistry Campaign announced a voluntary track-level labelling approach distinguishing between:
AI-Generated
and
AI-Assisted.
ARIA welcomed the standard, saying it would help listeners understand how music was made.
This creates an important distinction between labelling and eligibility.
A track can be AI-assisted and still be eligible for the ARIA Charts.
But a track that is primarily AI-generated can be excluded from the charts.
In other words:
Disclosure tells listeners what happened.
Chart rules determine what qualifies.
What Happens If ARIA Finds a Track Is Ineligible?
ARIA has given itself several options.
If it determines that a recording doesn’t meet the eligibility requirements, it can:
* Refuse to accept it for chart surveying
* Remove it from the charts
* Correct a previously published chart
* Adjust chart positions
* Withdraw accreditations
* Revoke or request the return of an ARIA #1 Award
And there is another important consequence:
A recording that is not eligible for the ARIA Charts is also not eligible for an ARIA Award.
This effectively connects chart eligibility with awards eligibility.
Musicians Will Have a Right to Challenge a Decision
ARIA’s new framework isn’t simply an automated “AI detector → instant removal” system.
Every release submitted to the ARIA survey now requires a declaration concerning generative AI use. If ARIA receives a credible concern about a recording, it can go back to the rights holder or representative for additional information.
If a musician disagrees with an exclusion, the artist or representative can provide evidence supporting the recording’s eligibility.
Disputes can then be referred to the ARIA Chart & Marketing Committee and subsequently to the ARIA Board.
That means documentation around how a recording was created could become increasingly important for musicians working with AI.
What This Means for Music Producers
For producers, the most important takeaway is that using AI doesn’t automatically make your music ineligible.
If you’re using AI for:
* Mastering
* Stem separation
* Sound design
* Production assistance
* Drum tools
* Instrument patches
* Minor creative enhancements
while the core musical performance remains human, ARIA’s current framework allows such use.
The bigger risk comes when AI replaces the primary human performance or creative contribution.
For example, an AI-generated vocal replacing the lead human performance is treated very differently from using AI to process or enhance a vocal that was actually performed by a musician.
For independent producers, this makes it increasingly important to understand exactly what your AI tools are doing rather than simply categorising everything as “AI-assisted.”
What This Means for Artists and Songwriters
The policy also reinforces something that has always mattered in music: documentation of creative contribution.
As AI becomes more integrated into production, artists may increasingly need to understand:
* Who wrote the composition?
* Who performed the recording?
* Which parts were generated by AI?
* Which parts were performed by humans?
* What AI tools were used?
* Were those tools authorised?
* What rights does the tool’s licence provide?
* Does the resulting recording comply with applicable copyright and related-rights laws?
This doesn’t mean every independent artist needs a legal department.
But it does mean that knowing what you are licensing, uploading and distributing is becoming more important.
The Streaming Industry Is Already Moving in the Same Direction
ARIA’s decision is part of a much broader platform-level response to synthetic music.
Deezer says its AI detection system has been operating since early 2025 and that it became the first streaming service to explicitly tag AI-generated music in June 2025. It also excludes detected AI-generated music from algorithmic recommendations and editorial playlists.
In June 2026, Deezer also launched a free AI music detector that users can use to check playlists across major streaming platforms. Deezer reported that 97% of listeners in its blind test could not distinguish AI-generated music from human-created music, while 80% said AI music should be clearly labelled.
This illustrates why transparency is becoming increasingly important: listeners may not always be able to tell the difference themselves.
The Grammys Are Taking a Similar Approach
ARIA isn’t the first major music institution to establish human-authorship rules around AI.
The Recording Academy says its AI rules, introduced in 2023, remain in effect: only human creators can be nominated for or win a Grammy. Works containing AI-generated material can still qualify when there is meaningful human authorship or performance appropriate to the category.
The Academy has also been advocating for broader protections around AI, including legislation addressing unauthorised digital replicas of artists’ voices, images and likenesses.
So across charts, awards and streaming platforms, a common principle is emerging:
AI can be a tool. But human creativity remains the benchmark for recognition.
What Does This Mean for the Future of Music?
ARIA’s new rules don’t answer every question about AI and music, and ARIA itself acknowledges that the policy will evolve as the technology and licensing landscape develops.
The most interesting part may therefore be what happens next.
As AI tools become better at generating vocals, instruments, arrangements and complete recordings, music platforms and industry organisations will need increasingly precise ways to distinguish between:
Human-made music
Human-made music with AI assistance
AI-generated music
Unauthorised AI-generated music
AI-generated music involved in streaming manipulation
Those categories could eventually affect not only chart eligibility, but also recommendations, royalties, awards, licensing, copyright, distribution and discoverability.
For musicians, the message isn’t necessarily “don’t use AI.”
It’s:
Know what you’re using. Know what you own. Know how the music was made. And keep the human contribution at the centre when your goal is recognition in systems built around human artistry.
What Musicians Should Take Away
ARIA’s decision provides a useful framework for musicians everywhere:
1. AI isn’t automatically the problem.
AI-assisted production remains eligible when the recording is substantially human-made.
2. Human performance matters.
AI-generated lead vocals or key instrumental performances can make a recording ineligible for the ARIA Charts.
3. Keep track of your AI usage.
Understanding what tools contributed to your recording can become important for disclosure and eligibility.
4. Don’t ignore licensing.
The wider IFPI framework places significant emphasis on authorised and lawful AI services, even though ARIA says it is not applying that particular test to chart eligibility yet.
5. Streaming manipulation is becoming a bigger issue.
AI-generated content combined with fraudulent streams is increasingly being targeted by platforms and chart organisations.
6. Your rights and metadata matter.
As the industry moves toward clearer AI labelling, accurate information about how a recording was created will become increasingly valuable.
Final Thoughts
ARIA’s new AI chart rules represent something bigger than a change to one country’s music charts.
They show the industry beginning to establish a line between AI as a creative tool and AI as a substitute for human artistry.
With tens of thousands of AI-generated tracks entering streaming platforms every day, the question is no longer whether AI will be part of the music ecosystem. It already is. The question is how the industry will recognise, label, monetise and regulate music created with it.
For musicians and producers, this makes one thing increasingly clear: the future of music may involve AI, but the value of human creativity, performance, ownership and rights is becoming more important, not less.
