South Australia is launching a royal commission into AI, and it wants the tech giants in the room
Premier Peter Malinauskas is spending $3 million to figure out how artificial intelligence should be regulated before the technology outpaces the rules meant to govern it.
Key points
- South Australia will launch a royal commission into artificial intelligence on 1 October 2025, with a final report due by 1 July 2026.
- Premier Peter Malinauskas expects the inquiry to cost approximately $3 million and involve three commissioners selected through an international search.
- The commission will examine AI's impact on education, health, public services, work and the electricity grid, but will not cover data centres.
- Malinauskas met with OpenAI and Anthropic, along with Apple, during a recent US trip before announcing the inquiry.
- UNSW AI Institute chief scientist Toby Walsh welcomed the move, saying it could push the federal government to act faster on AI policy.
South Australia is launching what is believed to be a first-of-its-kind government inquiry into artificial intelligence. Premier Peter Malinauskas announced the royal commission, a formal independent investigation with the power to compel evidence and make binding recommendations, saying AI left unchecked poses a "material risk to the way society operates".
The commission kicks off on 1 October and must hand its report to the government by 1 July 2026. Three commissioners will be appointed after an international search, with terms of reference published within the next four to six weeks.
What will it actually look at?
The inquiry will cover how AI affects schools, hospitals, public services, creative work and the job market. Data centres, the warehouse-scale buildings that AI companies need to run their systems, are explicitly out of scope. Malinauskas wants the focus on what AI does to people, not the physical infrastructure behind it.
He was blunt about the jobs question. "If we don't do this properly, we could see very significant changes in the way that people work and a more rapid destruction of demand in some areas of the labour market," he said. Workers in administration or customer service should read that sentence carefully.
The inquiry will hear from businesses, unions and technology companies, as well as academics. Malinauskas specifically wants AI firms to participate and said he met with representatives from OpenAI, Anthropic and Apple during his US trip.
Why South Australia specifically?
Australia has no single national approach to AI regulation. Some states have banned ChatGPT, a conversational AI assistant made by OpenAI, in schools outright. South Australia has not, and Malinauskas says he wants a considered framework rather than a blanket ban.
We covered Australia's federal AI policy push on 7 August 2026, and the South Australia commission looks designed to run alongside it rather than against it. Toby Walsh, chief scientist at the UNSW AI Institute, said the state inquiry could complement that federal work and push Canberra to move faster. "If this is a way of getting the [federal] government to move faster, I welcome the idea that South Australia can again help us get to a good place," he told reporters.
"There is no precedent in humankind for a new form of intelligence that is able to match, and soon exceed, the capabilities of human beings," Malinauskas said.
$3 million sounds like a lot. The premier called it "chicken feed" compared to the cost of getting AI policy wrong, and it's hard to argue with that logic.
Common questions
Does this affect how I can use AI tools in South Australia right now?
No. The commission is an inquiry, not a ban. Existing tools remain available while the review is underway, with any restrictions only possible after recommendations are made and adopted.
Will this commission change things nationally?
Possibly. Malinauskas explicitly said he hopes other state and federal governments follow the recommendations, and Walsh suggested it could accelerate federal action on AI regulation.
