A former judge and two global tech experts will lead a royal commission into artificial intelligence after the most "substantial technological shift" in two centuries.
The South Australian government on Thursday revealed the terms of reference for the state inquiry and its pick for chair, economist and former judge Iain Ross.
Premier Peter Malinauskas said it was impossible to put the "AI genie back in the bottle", conceding the fast-developing technology sometimes scared him.
"If we don't contemplate ... an AI future forcefully, there could be really severe consequences, as well as a big opportunity cost," he said.
Kate Pounder and Christopher Manning were selected as commissioners for their expertise in technology policy, innovation, computer science and machine learning.
Dr Ross said the inquiry would balance opportunities for enriching lives, improving productivity and lifting economic performance with an approach that mitigated risks and threats.
Professor Manning, a world-renowned AI expert, is the co-founder of Stanford's Institute for Human-Centred Artificial Intelligence and is a professor in machine learning at the private research university.
He would visit SA in early 2027 and his connections would provide the best possible expertise, Dr Ross said.
Ms Pounder, who was raised in the state, has worked with OpenAI and was the Tech Council of Australia's inaugural chief executive.
The commission could help people gain clarity about their future and provide support as people "make the most of this fundamental change", she said.
The evidence-based inquiry aimed to help SA industry be nationally and globally competitive and would study how the technology could be transformative to the economy, the terms of reference said.
It will also examine how SA can leverage the social benefits of AI while safeguarding people and society from risks and challenges.
"There has never been a more substantial technological shift ... over the course of the last two centuries and we're going to treat this seriously," the premier said.
"Anyone that's used (AI agents) ChatGPT or Claude has witnessed it themselves - it's extraordinary what you can do very quickly.
"But … when I think about my kids, I worry about things that aren't human-centred making judgments for us and what the implications of that could look like."
Labour market and workplace practices, including the mitigation of impacts on employment, will be examined, as will the education system's use of AI.
The commissioners will make recommendations about policy, legislative or structural reforms, with regard to the need for practical action in the short to medium term.
They will be required to consider the importance of technological innovation remaining human-centric, and current and potential impacts and opportunities of AI on systemic trust, civics and the democratic process.
The commission will co-ordinate public submissions, community forums, surveys, industry round tables and stakeholder consultations.
Australian Information Industry Association chief executive Elizabeth Whitelock said the inquiry could help establish the conditions for responsible AI adoption that strengthened the economy and improved lives.
"AI should help people do their jobs better, help businesses grow and make essential services easier to access," she said.
"People need confidence that AI is being used responsibly, with clear accountability and appropriate protections. Businesses also need practical guidance for how it can benefit them."
The commission is due to report to the government in mid-2027.
