ARIA makes human authorship an eligibility test for Australia's music charts

The people this rule reaches first are the ones who decide what to release. A label weighing an AI-assisted cover for the Australian market now has to guess whether ARIA will count the finished record as "substantially human-made," because nothing gets onto the chart without clearing that condition ^[2]. The reporting does not set a threshold for how much human input clears it.

A disclosure label and an eligibility test look similar in a policy document and behave nothing alike in the building. A label needs a field and a definition loose enough for an uploader to fill in. An eligibility test needs an adjudicator, and the adjudicator needs evidence they will accept from whoever supplies the record. ARIA has taken the second design ^[2].

A radio chart counts plays, and by that measure the AI-assisted cover of Madonna's "Like a Prayer" was the biggest record in the country in August ^[3]. Listeners revolted, and ARIA listened, in the words of the account of the episode ^[13]. The rule the association landed on removes wholly AI-generated songs ^[1], and the track that provoked it was AI-assisted.

Volume is what makes the adjudication expensive. More than a third of new uploads to Apple Music this year were entirely AI-created, according to figures cited by RTE ^[4]. Apple Music is not a chart compiler, but any provenance rule applied to a catalog gets tested against an intake like that.

Craig Anderson has spent 25 years restoring and remastering the analog masters behind catalogs from Green Day, My Chemical Romance, Red Hot Chili Peppers and Alanis Morissette, and founded the Los Angeles archive and encoding house Craigman Digital ^[8]. He said the vendors have already adjusted to the reaction. "AI companies today are looking at the outcry, and they're backpedaling quite a bit, and they're readjusting their message," Anderson said ^[9]. He pointed to advertising that opens by acknowledging fears over jobs and children's attention before turning to medical and scientific breakthroughs ^[15]. "They try to assure us that everything is going to be great, especially in those sectors, but they don't really address the level of employment, or the downturn in someone's education, where they don't have to think anymore," he said ^[10].

The audience evidence in the account is looser than the rule it is used to explain. Its summary says the 52% of Americans who are more concerned than excited about AI rises to 73% among 18-to-29-year-olds, while its body attaches the 73% to the separate question about job losses ^[11]. Both readings cannot describe the same measurement. The creative-sector figure is narrower and harder: one in ten of nearly 900 Creative Boom respondents called AI's effect on their field positive, which is about 90 people ^[7][12].

For anyone shipping generative features next to a ranked surface, the useful split is between surfaces that ration distribution and surfaces that only describe it. A chart slot or an editorial playlist position is scarce, so a provenance condition on it is an eligibility rule and needs a named adjudicator. A badge on a track page rations nothing, and a self-declared field will carry it. The practical test is whether you can name the artifact a submitter would hand over: the stems, the session file, a signed statement from the producer. If nobody can name it, the policy only describes the record, and the ranking stays exactly as it was.