ARIA Bars Fully AI-Generated Music From Australian Charts
The Australian Recording Industry Association (ARIA) has ruled that tracks entirely generated by artificial intelligence are no longer eligible for Australia's music charts. The policy, announced August 25, 2026, takes effect the Friday following the announcement. Recordings that fail chart eligibility will also be barred from winning an ARIA Award. Fortune
The distinction ARIA draws is between tracks produced entirely by AI and those where a human creator used generative AI as one tool among many. Under the updated Code of Practice, songs that incorporate generative AI remain eligible, provided the recording is "substantially human made" and raises no stream or chart manipulation concerns. ARIA's official Charts Code document, updated August 24, 2026, codifies this in Schedule 3: "Eligibility Rules for Particular ARIA Charts," under which recordings that are not substantially human created, including where artificial intelligence was used, are ineligible. ARIA ARIA Charts Code (PDF)
ARIA CEO Annabelle Herd framed the changes as a preservation of the charts' core function. She said the ARIA Charts "will always remain a transparent measurement of the music Australia consumes," and called on parties deciding what music is played and promoted to Australian audiences, particularly radio, "to support human artistry and implement similar changes." Engadget
The rules do not arrive in a vacuum. In the month before ARIA's announcement, a coalition of major labels, with support from Sony Music, Universal Music Group, and Warner Music Group, called for charting organizations worldwide to prohibit AI-generated "slop songs" from entering charts. ARIA's new policies also echo recommendations on AI use laid out in July by the International Federation of the Phonographic Industry (IFPI), the global trade body for the recorded music industry. The IFPI's guidance set out principles for the eligibility of recordings developed using AI in official music charts worldwide. Engadget IFPI
ARIA maintains a dedicated FAQ page addressing the changes, confirming that a recording made by humans who used generative AI is still eligible, contingent on the same "substantially human made" and anti-manipulation criteria. The FAQ does not specify a formal audit or verification mechanism for determining what constitutes "wholly" versus "substantially" AI-generated content. ARIA FAQs
The phrase "substantially human made" carries a substantial enforcement burden. Without a defined quantitative threshold, ARIA will need to adjudicate borderline cases on what is likely a case-by-case basis. A track where a human wrote lyrics and melody but used AI to generate the instrumental backing, or one where AI handled mastering and arrangement while a human performed the vocals, could produce reasonable disagreement about whether the "substantially human made" bar is met. The policy's second condition, that a track raise no "stream or chart manipulation concerns," is more tractable in that chart manipulation detection is already an established operational practice for charting bodies, though AI-generated content introduces new vectors, particularly volume-based flooding of streaming platforms.
Herd's call for radio and other promotion gatekeepers to adopt parallel standards broadens the policy's potential impact considerably. The ARIA Charts measure consumption, not airplay directly, but radio remains a significant driver of discovery and, by extension, streaming volume. If Australian radio networks align with ARIA's stance, the combined effect would extend the exclusion from charts into the promotional pipeline that feeds them.
The coordinated sequence here is worth noting: a major-label lobbying push, a global industry body's principles, and then a national charting body's binding rules, all within roughly two months. That compression suggests the recorded music industry has reached a working consensus on the parameters of AI's legitimate role in chart-eligible recordings, at least at the level of trade associations and major labels. Whether independent artists and smaller labels, who may rely more heavily on AI tools for cost reasons, share that consensus is a different question.
The broader context here is that ARIA's "substantially human made" standard is an early regulatory artifact in what will likely be a longer negotiation across creative industries over authorship thresholds for AI-assisted work. The music charts problem is, in structural terms, a ranking-integrity problem, and the tools and frameworks developed here, including definitional boundaries, verification mechanisms, and anti-manipulation controls, are likely to inform analogous decisions in adjacent domains where AI-generated content competes alongside human-created work in ranked or curated systems.
