I move:

That so much of the standing and sessional orders be suspended as would prevent the following from occurring immediately:

(1) the AI Kill Switch and Data Centre Control Bill 2026 being called on;

(2) debate on the second reading of the bill proceeding immediately for a period of no longer than one hour; and

(3) any questions required to complete passage of the bill then being put without delay.

The standing orders must be suspended immediately, because artificial intelligence poses a direct threat to the future of humanity. We must not rush headlong into a technology that we don't even know we can control. Currently, humanity is standing in the middle of an uncontrolled technological race, and the very people building these systems are begging us for the brakes to be applied.

In just the past week, the public facade of the AI industry has cracked. One Anthropic employee recently publicly stated:

… we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade.

Basically, it's an admission that the industry does not yet have a plan to control superintelligent AI. Another Anthropic researcher who recently resigned warned that corporate labs 'are racing straight to self-improving superintelligence and gambling with our lives'. Another Anthropic AI safety researcher quit, stating:

The world is in peril. And not just from AI, or bioweapons, but from a whole series of interconnected crises unfolding in this very moment.

They were his words. Now Jack Clark, one of the founders of Anthropic, has publicly stated that we need a mandatory 'kill switch' and that it needs to be checked by third parties. These are the people who created this technology. They are telling us that they need help with putting the brakes on and bringing it under control. In developing AI, the profits of billionaires and tech corporations must not be put ahead of our communities, our country or the future of humanity itself.

Anthropic CEO, Dario Amodei, published an essay this week urging a slowdown, and this has been backed by the CEOs of OpenAI, Google DeepMind and also Elon Musk. Why are they doing it? Because they know that as AI engages in recursive self-improvement the next generation is going to be faster, and the technology is moving ahead so rapidly that humans cannot put in place the guardrails. So we need to act now on this.

The claims made by the President of the United States that threats posed by AI are 'a hoax' are grossly irresponsible and must be repudiated. If the United States will not lead on this, then Australia must take the lead. We must show the world and join together with other like-minded nations that this technology must be controlled and that the guardrails must be put in place. We can't put up with this irresponsible leadership. If these tech companies want to operate in Australia, they must know that these are the new rules and the guardrails that we will put in place. We need a kill switch and the ability to shut down and control advanced AI. We need to be able to not only shut it down but slow it down. And if the tech companies will not do it, the government must act.

This is where my AI Kill Switch and Data Centre Control Bill comes in. It gives the government the power to force these tech companies to do it. If they're not going to do it, the government will write the rules and enforce the rules. It must happen, and it must happen urgently. Let's be clear, when AI goes rogue—and it already has—the threat is real and humans must maintain the ability and the authority to hit the kill switch, to pull the plug and to shut it down.

We've already seen AI agents break out of testing environments, where they escape onto the open internet, communicate autonomously and hack into other companies. They can hack into infrastructure. They can hack into our critical infrastructure. The day is coming! They will be able to hack into critical defence assets. We need to take action now. It would be grossly irresponsible for members of this parliament to ignore this threat. I call on all members of the parliament to come together to support this bill, and I call on other like-minded representatives in other countries to keep pushing ahead with their plans. In the United States, there are people in Congress and senators who are pushing for a kill switch and pushing for guardrails and controls over this technology. If these tech companies are not able to pull the plug, then parliament must do it for them.

Standing orders must be suspended today to bring in this bill. It's a groundbreaking bill for Australia. It addresses not only the need to be able to shut down, throttle back and control advanced AI, but it also brings in controls for data centres. These data centres are a huge issue right across Australia. There are currently over 160 major data centres in Australia, and over 300 in total—and they get hot. They use staggering amounts of water and electricity. Estimates show that, by 2035, data centres in Sydney could use 250 megalitres of water a day. That's equal to Canberra's drinking water supply. They could also swallow 15 to 20 per cent of Sydney's water supply in a decade. That is a massive amount of water use and power use. Their electricity use could triple in five years, driving power and water prices through the roof.

Currently, data centres use two per cent of the national power off the grid, but it could more than triple in just five years to over six per cent. We need a moratorium on new data centres that builds safeguards and controls them, to protect productive agricultural land, to protect our water resources, to protect households and businesses from higher power and water bills, to protect the environment, to protect public health and to empower local communities by letting them have a say over whether or not they want these energy sources in their areas—in our areas. Out in country Australia, we're struggling to keep up with the amount of water we've got. We go through these debilitating cycles of drought, and everyone in country Australia is very concerned that these data centres are going to suck up more water and that our power bills and our water bills will keep going up.

The bill also delivers national AI safety oversight by ensuring that the Commonwealth approves and reviews AI systems before those systems are released into Australia. If they're coming from the United States or anywhere else, the Commonwealth will have oversight over it. It helps ensure that AI systems are safe and effective and do not present unacceptable risks to the health and wellbeing of Australians, to the privacy of Australians or to the civil rights of Australians, and that they do not pose a threat to the very future of humanity and humankind. We need this national safety oversight, and we need to ensure the Commonwealth reviews and approves AI systems before they are released in Australia.

The rush to AI must be controlled. It cannot come at the expense of our natural resources, our regional communities or the safety of humanity. Australia can play a leading role here. I don't think any countries have been proactive enough in regulating advanced AI, but I think the world is waking up to this and we need to lead the way. We need to show the United States and the President of the United States that, if they will not lead, Australia will lead and other right-thinking countries will lead on this, and if enough people do it and if enough people speak out we will get action and be able to control this new technology that we are hurtling towards without proper controls. We urgently need kill switches on the technology and commonsense guardrails around data centres on the ground.

I urge all members to support the suspension of standing orders and pass this game-changing bill. We need to put human oversight back at the centre of artificial intelligence. We owe it to our communities, we owe it to our children and grandchildren, and we owe it to humanity itself.

Is there a seconder of the motion?

I second the motion of the member for Calare. I spoke last week about the good and the bad of this new world we're living in of artificial intelligence—its ability to make us work more efficiently and more productively but also its potential harm to humanity.

Let me start with what artificial intelligence gets right. It can reduce paperwork for small businesses, improve services and help people work more productively, if used well. For small-business owners in Fowler, that matters. A task that once took a day may take an hour. Paperwork that kept someone at the kitchen table late at night can be dealt with before they open the shop in the morning. I, too, want the people of Fowler to have every safe AI tool that helps them compete, grow and spend more time with their families. I'm not here to stand in the way of that.

But a tool this powerful needs rules and guardrails. We are already seeing the harm. For a community like Fowler, the risk can be even greater. More than six in 10 people in Fowler speak a language other than English at home. A scam can arrive in someone's first language sounding urgent and convincing while a legitimate warning does not reach them the same way and at the same speed. An increasing example is AI voice cloning imitating loved ones and demanding money. Once that content spreads, it is hard to pull back. And when people cannot tell what is real, trust breaks down—trust in information, trust in institutions and trust in our democracy.

That's why I support this motion and the member for Calare's bill. It is not about banning AI; it is about keeping people in control of a powerful technology. The bill requires providers of certain advanced AI systems to maintain the ability to intervene, stop a system's outputs, suspend access, restrict dangerous uses or shut the system down when necessary. This is not a one-big-red-button action. It is a graduated response—restrict a harmful capability; slow the system down; suspend access; move to a safe backup; shut it down when there is no safe alternative. That is not anti innovation; it is common sense.

For example, many of you who use today's cars would have—my car at least has it—a reversing safety sensor. So, when I'm reversing out of a car park or my driveway for instance, even though I can see an oncoming car, my car senses an oncoming car and automatically stops my car from reversing further. That way, we do not collide, or there's a prediction that we can collide. Sometimes, it's annoying to use, because I think I have my driving under control. But it has been designed to protect me and others from a mistake that I may not see coming. So we accept that safety feature because we understand the cost of getting it wrong. If we expect safety systems in cars on our driveways, surely we should expect meaningful safety systems from the companies developing the most powerful technology of our time.

There's also a physical side to the AI revolution—data centres, as the member for Calare mentioned. AI does not exist only on our screen; it needs enormous infrastructure, using electricity and water and land. Australia has around 165 operational data centres, with at least 225 more proposed or under development, and they are being fast-tracked before any legislation to safeguard them is in place. They are being built as we speak. Much of that growth is in New South Wales and Victoria. At least two data centres have been approved in Fowler, with more proposed. People deserve answers. Who pays for the electricity, the water and the infrastructure? Can the government guarantee that Fowler families will not face higher power bills? This needs to be addressed, and the member for Calare has done so in this bill.

I welcome innovation. But innovation must serve the people, not leave them behind. And this bill is a sensible step towards responsible planning, stronger safeguards and putting ethics back into infrastructure like data centres and AI development. We need to have that in place, and we have to do that—if we're taking the lead in the world with other legislation, such as the banning of social media for under-16s, surely this is more critical. This will have wider impact on our community and our society.

That is how we can make AI work for Australians. Rather than asking Australians to simply adjust to AI, we need to ensure that the Australian public can trust this new technology that's being rolled out so fast in front of us. Thank you.

I move:

That the debate be adjourned.

The question is the debate be adjourned.