Where Does the Human End? AI Creativity Is Forcing a Real Answer

A Chart Ban Nobody Saw Coming

Late August brought a policy decision that caught a lot of AI creators off guard. The Australian Recording Industry Association (ARIA) announced on August 24 that it was updating its rules to distinguish between music generated primarily by AI and music that uses the technology as a supporting tool, with the new rules taking effect with the chart dated August 31, 2026.

In plain terms: music produced with generative AI can only appear on the ARIA charts if it was "substantially human made" and "raises no stream or chart manipulation concerns," and wholly AI tracks are also banned from eligibility for ARIA's Music Awards.

This didn't come out of nowhere. The crackdown comes as a variation of Madonna's pop hit "Like a Prayer" created by an Australian producer using AI-generated vocals and drums spent 16 weeks in the Australian top 20, peaking at number two on the Australian singles chart in May. The song sparked real debate — and apparently real industry anxiety.

What makes this decision interesting for AI creators isn't just the policy itself. It's the question underneath it: what actually counts as a human creative act?

"Substantially Human Made" Is Harder to Define Than It Sounds

ARIA's answer is specific but still leaves a lot of room for interpretation. Tracks can be eligible only if "humans wrote the song and performed the lead vocal and the primary instruments," among other requirements. AI-assisted mastering will also remain allowed.

So you can use AI to clean up your mix. You can't use it to generate the melody, sing the hook, or play the lead. The line ARIA is drawing is essentially: the expressive core must come from a human.

That's a philosophically interesting threshold. A producer who programs every MIDI note by hand but uses an AI vocal synthesizer to "perform" their melody would likely be out. A singer-songwriter who records with a live band but uses AI mastering is fine. The instrument matters less than the performance.

ARIA's decision comes as AI's presence in the music industry becomes harder to ignore — a new study by SubmitHub found that almost 40 percent of new music released globally in July 2026 used AI in its creation. Of that, 23.2 percent were fully AI-generated, while another 15.3 percent contained AI-generated audio that had been "modified or processed" by humans. Those numbers are staggering, and they explain why the industry felt it needed to act.

This Isn't Just About Music

The ARIA decision is the most visible recent move, but the tension it reflects runs through every creative discipline where AI tools have become mainstream — visual art, writing, video, sound design.

For the AI creator community, the uncomfortable question is this: is AI a brush, or is it the painter?

The "brush" framing has always been how most AI creators describe their own work. They direct. They prompt. They curate from dozens of generations. They iterate. They bring taste and intention. The AI is the instrument; the human is the artist. This is a completely defensible position — and for many creators working this way, it's genuinely accurate.

But the "painter" concern is legitimate too. When a tool can generate a finished, commercially competitive track with minimal human input, something about the traditional creative equation has shifted. As ARIA put it: "Australian artists are competing for attention in the most crowded market in history, and an AI-generated recording taking a chart place matters to the people who create human artistry."

This is the economic argument, not just the aesthetic one. Chart positions affect streaming payouts, booking fees, label interest, and career trajectories. If AI-generated tracks can compete for the same slots, they're not just an aesthetic curiosity — they're a structural pressure on human livelihoods.

The Disclosure Question Is Probably More Important

One thing that seems clear from the ARIA debate: the problem isn't just AI content, it's unlabeled AI content. SubmitHub founder Jason Grishkoff argued that "the path forward for AI music is disclosure — people should be able to decide for themselves whether they want to engage with AI-generated music," and Spotify is now taking steps to make AI-generated artists easier for its listeners to spot.

That framing reframes the whole debate. If listeners knowingly choose to stream AI-generated music, whose problem is that? If they're unaware, that's a different ethical situation entirely.

ARIA has said it will evaluate recordings based on the level of human involvement, whether AI tools were used lawfully, and whether there are concerns about chart manipulation — which suggests the organization is treating this as an ongoing judgment call rather than a hard technical filter. That's probably realistic. A binary AI/not-AI distinction is technically impossible to enforce at scale; a principled, case-by-case approach with clear criteria is more honest about the complexity involved.

What This Means If You're Creating with AI

If you're an AI creator — whether you're making music, images, video, or writing — the ARIA ruling is a useful mirror, even if you're not in the music industry.

First, think about disclosure. Audiences and platforms are increasingly going to expect it, and getting ahead of that norm is better than being caught in a retroactive credibility problem. Labeling your work as AI-assisted (or AI-generated, where that's accurate) isn't an admission of lesser value — it's an accurate description of the process.

Second, think about where the human actually is in your work. Not as a defensive exercise, but as a creative one. The AI creator community has always been strongest when it's been honest about the fact that curation, taste, direction, and refinement are real skills — and that prompting well at scale takes genuine creative intelligence.

Third, recognize that industry norms are being written right now. ARIA's move comes amid a wider dispute between creative industries and AI companies, many of which have built their technology using copyrighted material without artists' permission. These aren't separate conversations — the training data debate, the chart eligibility debate, and the compensation debate are all part of the same larger negotiation over who owns the creative economy in an AI-native world.

ARIA drew a line. More industries will draw their own. The creators who will navigate this best are the ones who've already thought clearly about where they actually stand — not just what the rules require, but what they actually believe about their own work.

The question isn't going away. It's getting more specific.