Australia's ARIA draws a line on AI music, as data-centre electricity use is projected to soar nearly 600% by 2036

Australia's main album and singles chart is moving to keep fully machine-made songs off its rankings, on the same news cycle that surfaced a projection of national data-centre electricity use rising close to 600% by 2036.

On 25 August 2026, ARIA, the body that administers Australia's flagship charts, said it was updating eligibility rules to draw a line between artists who use artificial intelligence as a tool and songs created largely by machines, according to Deutsche Welle's English-language wire. The decision landed the same week that the Polymarket-curated news feed flagged a projection that data-centre electricity consumption will soar close to 600% over the coming decade.

Read together, the two items sketch a country in the middle of a compute build-out, trying to decide in real time which cultural outputs are human enough to count, and how much power the machines behind them should be allowed to draw.

A chart with a conscience

Deutsche Welle's reporting frames the ARIA move as the body drawing a line between AI as a tool and songs that are largely machine-made. The cited DW item does not, however, specify the exact eligibility criteria, the carve-outs, or the date the new rule takes effect. The available source material points to a direction rather than a fully articulated policy text.

Monexus analysis: the practical effect in week one is likely small, but the symbolic weight is larger. ARIA chart positions shape radio play, label marketing spend and the streaming-platform editorial lift that comes with a top-ten placement. Whether ARIA has settled the edge cases (a producer who fine-tunes a model on their own vocals, a session singer mimicking a prior hit, a release credited to a synthetic persona) is not addressed by the cited items.

A grid under pressure

The same news cycle carried a separate data point. The Polymarket-curated feed flagged a projection that Australian data-centre electricity use will soar close to 600% by 2036. That figure is presented in the cited item as a projection attributed to Australia; the cited items do not specify which agency produced it, what baseline year the 600% growth is measured against, or how the figure compares with earlier official estimates. The underlying source for the projection has not been independently verified in this article.

That uncertainty matters. A near-sevenfold increase in data-centre electricity consumption (the arithmetic of a roughly 600% rise) does not land on an empty grid. Australia has spent the past year pitching itself as a stable home for hyperscale compute in a region short on it; whether the cited projection originates with a federal agency, a state utility, or a private forecaster summarised in a news feed is not stated.

The labour question that does not name itself

ARIA's framing leans on a notion of human authorship that the cited material presents as the chart's organising principle. The edge cases the rule must still adjudicate go unstated in the available reporting. A producer who uses a fine-tuned model trained on their own ten years of vocals has arguably done more artistic work than a session singer hired to mimic a previous hit. The chart rule, as described by DW, does not visibly settle that question.

This matters because the chart is, among other things, a paymaster. Royalties flow through chart position to songwriters, performers and publishers. If AI tooling compresses production teams from five people to two, the chart still counts the song, but the pay cheque spreads across fewer bank accounts. Whether ARIA's new rule changes that arithmetic is not specified in the source items.

What the next twelve months actually look like

The cleanest reading the cited items support is that Australia is legislating the cultural edges of the AI transition while building the physical substrate for it. The chart will police authorship. The grid will absorb the load. Neither decision is more or less important than the other; the cited items treat them as two halves of the same news cycle.

Monexus analysis: a separate Polymarket-curated item, dated 24 August 2026, notes a U.S. appeals court considering the removal of a federal judge after his court issued an AI-generated order containing fabricated material and other serious errors. Read alongside the ARIA story, the two items point to the same underlying question landing in different jurisdictions: when an AI system produces a piece of text or music, who is responsible for its content, and what institutional rule should bind that responsibility? The cited items do not name the judge or the appeals court.

Desk note: Monexus paired an ARIA chart-rules item from Deutsche Welle's wire with a data-centre electricity-use projection surfaced by the Polymarket-curated news feed, plus a separate Polymarket-curated item on a U.S. federal judge and an AI-generated court order. The cited items do not specify the exact ARIA eligibility criteria, the originating agency for the near-600% data-centre electricity-use figure, its baseline year, or the names of the federal judge and appeals court involved. This article has not independently verified the data-centre projection against a primary Australian government source.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://www.dw.com/en/australia-s-music-chart-bars-fully-ai-generated-music/a-78491743?maca=en-rss-en-all-1573-rdf
• https://x.com/Polymarket/status/2092112483217518770
• https://x.com/Polymarket/status/2092016586366320881