Australia’s music industry is moving to keep fully synthetic tracks out of its official rankings, with new rules barring AI-generated music that is not substantially created by humans from the ARIA Charts. The Australian Recording Industry Association says recordings can still use artificial intelligence in a supporting role, but human creators must remain responsible for the central creative elements of the work.
The policy takes effect with the charts released Friday, Aug. 28, and gives ARIA authority to remove recordings it determines are ineligible, change chart positions and revoke chart awards. The change follows growing concern across the music business about whether increasingly capable generative-AI systems can compete directly with artists whose work, voices and performances may also have helped train those systems.
AI-Generated Music Must Be Substantially Human-Made
ARIA’s new definition distinguishes between “AI-Generated” and “AI-Assisted” recordings. Under its published guidance, a track can be considered AI-generated when generative AI produces the entirety or primary portion of its creative elements, including a lead vocal, a key instrumental performance or music produced largely from prompts.
AI-assisted tracks can remain eligible when humans substantially create the recording and generative AI is used only for some expressive elements. ARIA says human performers must provide the lead vocal and primary instrumentation under that category, establishing a practical dividing line between technology used as a production aid and technology functioning as the principal performer.
That distinction is important because an outright ban on all AI use would affect far more than fully synthetic music. Producers increasingly rely on software that can assist with editing, mastering, effects, composition and other parts of the creative process, and ARIA’s policy leaves room for those tools as long as meaningful human authorship remains at the center of the recording.
ARIA’s updated code also requires that qualifying music avoid stream or chart manipulation concerns. The broader policy is designed to preserve room for human-led experimentation with AI while drawing a firmer line around recordings in which generative systems provide the primary creative performance.
A Chart-Topping Madonna Cover Helped Force the Issue
The rule change arrives after Australian producer and DJ Josh Fawaz drew attention with a cover of Madonna’s “Like a Prayer.” The recording became the most-played song on Australian radio and reached No. 4 on two ARIA charts in July.
Fawaz initially described AI as a tool in his music-making process. He later added credits on Spotify identifying generative-AI vocals and AI drums on the track, making the song a prominent example of the increasingly blurry boundary between conventional production and machine-generated performance.
The controversy exposed a problem music charts are likely to face more frequently as generative systems improve. A song can attract real listeners, streams and radio play even when some of the performances generating that demand do not come from human musicians.
Under ARIA’s new framework, a track using AI-generated lead vocals or key instrumental performances could fail the substantially-human-made standard. An ARIA spokesperson told ABC that rules based on the global industry framework would likely exclude those types of recordings, although individual eligibility decisions remain subject to the association’s process.
Music Industry Pushes for AI Transparency
ARIA’s decision is part of a broader effort by the recording industry to establish clearer rules around artificial intelligence before synthetic music becomes indistinguishable from conventional releases for many listeners. The association says its new approach aligns with principles advanced across the global recording industry to distinguish substantially human-made recordings from wholly AI-generated material.
Streaming platforms are also experimenting with transparency measures. Spotify has introduced AI-related credits that allow artists to disclose the use of generative systems, while also taking steps against large volumes of low-quality or manipulative content entering its catalog.
The dispute goes beyond labeling. Record labels, performers and rights holders have argued that some generative-AI systems were developed using copyrighted music without sufficient permission, raising questions about whether companies should be allowed to build commercial models from creative works and then use those models to generate competing products.
That debate mirrors transparency battles elsewhere in artificial intelligence. American News Brief recently examined how Anthropic is adding machine-readable marks to Claude-generated content, another example of technology companies and regulators attempting to make synthetic content easier to identify as AI spreads across text, images, audio and video.
ARIA Charts Become a Test for Human Authorship
ARIA’s approach does not attempt to stop musicians from using artificial intelligence. Instead, it establishes a condition for participation in one of Australia’s most visible measures of commercial music success: the recording must still be substantially the product of human creativity.
That may prove easier to state than to enforce. Modern music production can involve dozens of digital processes, and determining when AI assistance becomes AI authorship could require detailed disclosures about vocals, instruments, composition and production.
ARIA’s code gives the organization mechanisms to act when a recording fails its eligibility rules, including removing material from rankings and revising chart positions. Artists and representatives can also challenge eligibility decisions, providing a process for disputes when the boundary is not obvious.
ARIA CEO Annabelle Herd has urged organizations responsible for promoting music to Australian audiences to support human artistry and consider comparable changes to their own rules. That could put pressure on radio stations, streaming services and other gatekeepers to establish similar standards.
AI Music Rules Could Spread Beyond Australia
Australia is not acting in isolation. The global recording industry has been moving toward rules intended to distinguish human-created recordings from wholly AI-generated material, with more than 20 official chart programs adopting similar principles.
That makes the Australian policy a potentially important test case. If chart operators can consistently determine which recordings remain meaningfully human-made, the model could offer the music industry a middle path between banning AI technology and allowing unlimited synthetic production to compete directly with human performers.
The larger economic question is who should benefit as generative technology lowers the cost of producing music. Artificial intelligence can give independent creators access to capabilities that once required expensive studios or hired performers, but the same technology can also produce enormous volumes of tracks at marginal cost.
For human artists, the concern is no longer theoretical. Synthetic voices, instruments and entire compositions are already capable of reaching streaming services, radio stations and commercial charts, forcing the industry to decide what exactly a chart position is supposed to reward.
ARIA has now given its answer: technology can assist the artist, but it cannot replace the artist and still expect equal treatment on Australia’s official charts. The practical challenge will be enforcing that distinction consistently as the tools become more sophisticated.
