Kate Chaney (Curtin, Independent) Share this | Link to this | Hansard source

Artificial intelligence is already transforming our lives, from how we work and learn to how we make decisions and solve problems. It brings enormous potential to improve health outcomes and research, streamline government services and drive productivity. But it also poses serious risks: disinformation, bias, sexual exploitation and a loss of privacy. Getting the policy landscape right is going to be essential to ensure we maximise the good and limit the bad. But AI is moving fast and the government is already falling behind, so what must we do?

First, we need a national AI taskforce and a clear plan. As the Tech Policy Design Institute has said, too often our choices are framed as binary: hit the brakes or slam the accelerator. But we can, and must, take hold of the steering wheel. To take hold of the steering wheel, we need a taskforce that's capable of responding rapidly and holistically to AI development. To regulate in a timely way, this taskforce needs access to key policymakers and regulators. We need a coordinated plan to provide a clear direction to industry and ensure we're not working in silos.

Second, we need an overarching legislative approach. Whether it's a standalone AI act or extending existing regulation, the goal is the same: support innovation while limiting harm. We must avoid getting bogged down in structure and focus on outcomes: safe, responsible and inclusive AI.

Third, we must address the high-risk uses of AI that will inevitably escape an overarching legislative approach. Last month I introduced a bill to criminalise the distribution and training of AI models designed to produce child sexual abuse material. We'll need to continually and rapidly stamp out horrific and unjustifiable uses of AI like this.

Fourth, we need to invest in our AI capabilities, which means investing in our people, businesses, schools, startups and research organisations. Already we've spent so much time debating regulation and whether we're going to slam on the brakes or hit the accelerator that we've forgotten to put fuel in the car. Australia's investment in AI has been modest compared to similar countries. Canada just announced $2.4 billion in funds for AI research, which is 17 times what the Australian government has invested over the past five years. But we have enormous potential. Just look at Maincode's Matilda, a world-leading AI model built right here in Australia. Investment in our AI capability goes beyond funding startups and infrastructure like data centres; it's primarily about our people. That means supporting people in small businesses, schools and across our communities to deploy AI effectively and safely. It means developing world-leading expertise in our universities and research institutions and creating pathways for global AI talent to contribute here.

AI presents an exciting frontier for this country, but we need to get a move on to ensure we can make the most of the opportunity in a safe way.