South Australia launches AI royal commission amid caution

The South Australian government has launched a Royal Commission into artificial intelligence, drawing a mixed but largely cautious response from industry and legal experts.

The move is one of the most comprehensive inquiries into AI undertaken by an Australian state government.

The commission will examine AI's impact on jobs, safety, regulation and public services. It will also consider how Australia positions itself in the global market for advanced digital technologies.

Industry leaders in South Australia see both opportunity and risk, arguing the inquiry must protect the public without undermining local innovation.

Tom Cawley, mining sector lead at AI Accelerator Cooperative Research Centre and executive chair and founder of MaxMine, said the state has an interest in shaping AI policy from the perspective of businesses already building and deploying the technology locally.

"As a South Australian company that has developed our own AI product here in the state, we welcome the Royal Commission, provided it strikes the right balance.

"We still don't know exactly how AI will play out. Will it destroy jobs, or drive a leap in productivity that kicks off another phase of economic growth? South Australia must make sure it captures the advantages of AI rather than watching them accrue elsewhere.

"Australia should be leading the world in applying AI to critical industries such as mining, where the upside for productivity, safety and efficiency is huge. This means going into a Royal Commission with our eyes and minds open, with the goal of managing the risks while ensuring we maximise the opportunity."

South Australia hosts a growing cluster of AI-focused firms linked to mining, defence and advanced manufacturing. Many say clearer regulatory settings will influence where global businesses place investment and research operations.

Legal experts have raised broader questions about how the Royal Commission frames its work, noting that the design of the inquiry will determine whether it produces practical rules and standards or only high-level observations.

Laina Chan, practising barrister and founder and chief executive officer of MiAI Law, said there is a need for a structured national conversation about how AI interacts with professional judgement and accountability.

"An inquiry of this kind is needed. Whether a Royal Commission is ultimately the best mechanism is open to debate, but it is a good way to get the ball rolling on some very important questions that Australia needs to confront now.

"The greatest misconception about AI is that a sufficiently capable language model can replace domain expertise. It cannot.

"Lawyers and doctors spend years developing judgement, not simply accumulating information. They learn how to distinguish relevant from irrelevant authority or evidence, reconcile conflicting material, recognise exceptions, identify uncertainty and understand when apparently similar cases are legally or clinically different. That judgement is the product of years of structured education, supervised practice and professional accountability.

"A language model is fundamentally different. It is trained on enormous volumes of material of vastly varying quality and produces remarkably fluent and persuasive answers. But fluency should never be mistaken for expertise.

"That distinction is already becoming critically important because people are relying on general-purpose LLMs for legal and medical advice. When someone treats a plausible AI-generated answer as though it were the advice of a trained lawyer or doctor, the consequences can be dire and potentially catastrophic.

"That is one of the issues I would want a Royal Commission to examine carefully: where AI can safely augment human expertise, where human professional judgement must remain central, and what safeguards are needed when AI is used in circumstances where people may rely on its answers.

"The question is not whether Australia should embrace AI. We should. The challenge is making sure we understand the difference between an extraordinarily capable language system and genuine professional expertise.

"If a Royal Commission can move that discussion beyond broad claims that AI is either wonderful or dangerous, and towards practical answers about reliance, accountability and high-consequence use, then it could be very worthwhile."