Aaron Violi (Casey, Liberal Party, Shadow Minister for the Digital Economy) Share this | Link to this | Hansard source

I move:

That this House:

(1) acknowledges the rapid changes artificial intelligence (AI) is driving across Australian workplaces, including:

(a) automation of routine tasks;

(b) augmenting jobs and increasing productivity;

(c) creating new jobs and skill sets; and

(d) workplace restructuring and business changes;

(2) notes that the Government has not developed an AI transition plan for Australian workers adversely affected by workplace restructuring and business changes, meaning the Government is silent on how Australians who lose their job because of AI will be supported or transitioned to another industry;

(3) recognises that conflict and tension between Government ministers and members is causing:

(a) AI policy inertia and delays;

(b) uncertainty for business to invest in AI technology; and

(c) serious unknowns for Australian workers impacted by potential workplace changes;

(4) further notes that after the Government won the 2025 election, it suddenly scrapped an expert AI advisory body after spending 15 months and $188,000 finding experts to join it; and

(5) calls upon the Government to resolve its internal conflicts and act now so Australian businesses, workers, and investors have certainty and clarity over Australia's AI policy direction.

There's a great saying: if you don't have a destination, any path will get you there. Unfortunately, that saying sums up the Albanese Labor government when it comes to artificial intelligence—no destination, no plan. Even worse, they had a destination that wasn't a destination that Minister Husic was taking them under, and then when Minister Husic got rolled the destination changed again. So now industry is left with a world with no certainty, no destination and changing priorities depending on the minister's whim. They rightly wonder what happens. To use the words of then minister Husic and still the member for Chifley, if the factional assassins come for Minister Ayres, does Assistant Minister Charlton get promoted? Does he take it in a different direction? Industry has no certainty. They've seen this play out over four years under this government.

Let's understand, when it comes to artificial intelligence, how important this transformation is for our society. While the technology is new, the transformation is not unprecedented. Society has gone through big technological changes before—the printing press, electricity, the internet, smartphones and social media just to name a few. So it's a question for government of how they capture the upside but make sure that they mitigate the downside, and providing that certainty not just for businesses to invest but also for workers is so crucial.

We have seen silence from this government when it comes to workers and how they will support them. Undoubtedly, they are waiting to get their marching orders from the unions on how they will look after communities when it comes to AI. We saw this play out in the last term of parliament when it came to industrial relations. We saw outsourcing to the unions resulting in productivity—less than a five per cent—decline under this government, productivity going backwards significantly. But they needed an action plan through microcredentials, through training people today. In these ways they can protect people, because we are seeing jobs being lost today.

Those opposite will look to spin and look to talk about how they actually have a plan and they are doing things when it comes to AI. I have no doubt they're going to reference the plan that they announced in December, and I look forward to hearing the references. They won't talk about how, within that plan, there was no framework. There were no real actions. There was talk about more conversations. Even worse than that, those opposite will potentially talk about the AI accelerator fund. They will talk about the money they're spending in the AI accelerator fund and how that is going to help businesses invest in artificial intelligence and our community take advantage of it. But that is a false claim by this government, because that fund does not provide money to industry until 2028. The one plan, the one idea, this government has to support artificial intelligence is to provide industry with funding in 2028—two years away. That is a lifetime when it comes to investment decisions and to artificial intelligence. We are seeing the changes move so quickly.

We know that this government has no plan and that all its ideas are talk. We know how bad it is because the member for Chifley—a member of this government and the former minister in this area—has consistently and publicly criticised this government when it comes to failures in artificial intelligence. The Prime Minister talks about how unified they are, how supportive they are, but what does it say that a member of his own government is prepared to take the risk to publicly come out and criticise this government? We all know what happens to those on the ALP side when they criticise the government. Senator Payman is a living example of that.

This government's own members admit it has failed when it comes to AI, and this is a technological revolution that will impact every Australian. It is one that we need to get right. There are so many opportunities to drive productivity, to drive economic growth and to improve our communities, but there is downside risk. We cannot afford to have a government that is asleep at the wheel, sending us in two different directions, then flipping those directions and flipping them again—providing no certainty for businesses, no certainty for the community and no certainty for the workers of Australia.

Mary Aldred (Monash, Liberal Party) Share this | Link to this | Hansard source

Is there a seconder for the motion?

Simon Kennedy (Cook, Liberal Party, Shadow Assistant Minister to the Leader of the Opposition) Share this | Link to this | Hansard source

I second the shadow minister's motion and I reserve my right to speak.

Jo Briskey (Maribyrnong, Australian Labor Party) Share this | Link to this | Hansard source

Artificial intelligence is bringing a massive shift to our workplaces, and this Labor government is working to ensure everyday Australians aren't left behind. Looking at the motion put forward today, you'd be forgiven for thinking that those opposite have suddenly become the great defenders of the Australian worker. Frankly, it's a bit rich coming from a party with a long and miserable track record of attacking workers' conditions. They want us to forget that they spent the last decade fighting against better pay, trying to strip away essential protections and literally cheering when our local auto manufacturing industry closed its doors. Now they want to be seen as the saviours of workers facing technological change. Give me a break!

We all know what their real approach to technology and industry looks like. Just a few weeks ago, Senator Bragg let the cat out of the bag when he was asked for his views on AI. His answer was simply that the government just needs to get out of the way. That is completely at odds with what our government believes. We back working people. We know that by getting everyone around the table—unions, employers and the training sector—we can make sure workers actually get ahead.

Despite those opposite having no road map on AI, they are making an absurd claim that we have no plan for workers whose jobs are changing, and that is simply not true. Helping people navigate these shifts doesn't mean leaving them to fend for themselves, as Senator Bragg suggested we do. It requires genuine investment in lifelong training—something those opposite have historically undermined through repeated cuts to skills and training. It requires making workplace laws stronger, which, again, historically, those opposite have banded against. It means ensuring the enormous economic gains from this technology flow to working people and not be concentrated with the few, as those opposite have consistently preferred.

In other words, the work needed to protect workers goes right to the heart of Labor's bread-and-butter priorities. It's exactly why we've invested heavily in education and why we are rolling out fee-free TAFE. Through our massive $12.6 billion National Skills Agreement, we are supercharging the VET sector and making sure Australians have the hands-on digital skills that they need for tomorrow. This is what real support looks like—not the empty rhetoric of an opposition that will cut funding to skills and training the second that they get the opportunity.

The coalition also tries to claim that there is policy inertia and confusion on this side of the House. Again, I think they're living in their own fantasy land, not in reality. Our National AI Plan is a crystal-clear road map to make sure these new tools build a fairer country. The economic upside is huge, and we're looking to inject up to $200 billion into our economy every year, generating 150,000 brand-new jobs by the end of the decade. We are focused on grabbing on to those economic opportunities, making sure that the benefits reach every postcode and putting the right guidelines in place to keep people safe. Instead of the opposition's bizarre claim that we've scrapped the expertise—we're actually putting $29.9 million on the table to build a permanent, dedicated AI safety institute. Getting up and running early next year, this institute will give our regulators the teeth and the expert advice that they need to stay ahead of the curve.

Just today, we announced a new, strict five-step framework for tech giants wanting to build data centres and use artificial intelligence in Australia. Under this framework, these companies must help pay for the renewable energy infrastructure and grid upgrades, with the aim of ensuring costs aren't passed on to households and businesses. These tech companies will also need to use water responsibly and make their computing power available to local businesses. Across our government, ministers are moving in the exact same direction. From workplace relations to education, we are heavily coordinating the rollout.

We're doing all of this because we know that AI, when implemented with a real understanding of our community's needs, can be genuinely life changing. Earlier this month, I visited Dementia Australia's pop-up here at parliament, where they showed how they train healthcare workers in AI through immersive simulations to better support people living with dementia. It's a practical example of this technology at its best. Artificial intelligence is reshaping the world, but our government's approach is grounded in our Labor values of fairness, inclusion and opportunity. While those opposite often lose sight of these principles, Labor is committed to ensuring AI enhances workers' skills and never simply replaces them.

Simon Kennedy (Cook, Liberal Party, Shadow Assistant Minister to the Leader of the Opposition) Share this | Link to this | Hansard source

The AI Safety Institute is a good idea. Do you know how we know it's a good idea? It was first announced on 18 June in 2021. The member of the government said that the coalition did not have an AI plan. That's actually incorrect. On 18 June 2021, the Australian government, then led by Prime Minister Morrison, published Australia's AIaction plan. That was one of the first action plans anywhere in the world. We had to wait years for the Albanese government to come out with something, and, when it came out with its plan, it was pretty much a carbon copy of what we had published only three years earlier. Yes, we suggested establishing a national AI institute. Yes, we looked at strategies for digital skills. Yes, we looked at AI ethics.

Australia has three massive problems. We have chronically low productivity, high inflation and low economic growth. Yet we have a solution staring us straight in the face: artificial intelligence. Artificial intelligence is perhaps the biggest lever in the world right now for productivity, yet Australia is not doing enough. Artificial intelligence is a race, and, sadly, it's a race Australia is losing. Right now, the US, China and many other middle powers are powering ahead. They're actually securing chips, securing energy and securing data supplies for their countries while Australia idly does next to nothing. We've empowered unions to promote AI. If you look around the world at any of these leading countries—be it Korea, the US or China—they do not have industrial unions driving their AI policy. That's what we have here at home. What could the government be better focused on? Reliable, cheap energy that will power this nation and power an AI world.

The truth is, Minister Bowen drew up our energy grid plan before the AI world and before blockchain was using an enormous amount of power. How do we know this? The energy market simply does not function any more. Energy prices are up 40 per cent in a retail sense. Many commercial areas in my area—manufacturers in Cook, Caringbah and Taren Point—have energy prices that are up almost 100 per cent. And what are they having to do to accommodate these energy prices? They're firing workers. FJP Manufacturing in Caringbah used to have seven apprentices. It's down to one. Energy costs are spiking, revenues are declining and there is no way to claw back these increased costs.

In AI, data centres and big tech, we are facing the same energy crunch. We have global companies looking for where they are going to invest tens of billions of dollars. What do they want to see? They want to see the Australian government help secure chips, provide security of energy supply, provide low-cost energy and fast-track permitting for data centres and the infrastructure that will build the knowledge revolution of this country. But instead we have crickets.

Yes, we have an AI institute, and I commend the government on that. That is smart and clever policy. But this isn't just about putting up an AI safety institute; this is about winning the race. This is akin to COVID. We are competing against the US, China and the whole developed world for hundreds of billions of dollars of capital and for jobs, and they won't naturally choose Australia. We are a smaller middle power. If we do nothing, this investment will go to the US, to China and to Russia—to these large superpowers.

This government needs to lead. They need to lead with energy policy. They need to lead with securing chips. They need to lead with fast-tracking permitting for data centres. They need to lead with catalysing private capital to invest in Australia. They need to work out copyright solutions so we can not just rely on inference modelling but build frontier modelling right here in Australia, right now.

The coalition is willing to partner with the government on this. If they come to the table and show how we will catalyse lower-cost energy, cheaper chips, secure energy supply and fast-tracked permitting, we are willing to work with them. But right now we are not seeing that commitment from the Labor government, and, sadly, Australia's economic growth, Australia's next generation and the wealth of the Australian households will all be poorer for this.

Tom French (Moore, Australian Labor Party) Share this | Link to this | Hansard source

I rise to respond to the motion moved by the member for Casey. At the outset, it is necessary to address the premise of the motion. It presents itself as a defence of Australian workers in the face of technological change. That is a curious position from those who have consistently opposed stronger workplace laws, argued against lifting wages and previously supported the exit of industries from this country. The House has been asked to accept that those who say the government should get out of the way are now the principal advocates for workers affected by artificial intelligence. That is not a serious policy position. It is a contradiction.

Artificial intelligence is shaping the global economy. It is changing how Australians work, learn and connect. The issue before the House is how that change is managed and in whose interests. The government's position is clear: we do not accept that the answer is to step aside and hope the market resolves these issues. Technological change must be shaped so that its benefits are broadly shared and its risks are properly managed.

That is why the government has developed the National AI Plan. It is a coordinated framework to ensure AI contributes to a stronger and fairer economy. It is anchored in three objectives: capturing the opportunity, spreading the benefits and keeping Australians safe. Estimates indicate that AI could contribute up to $200 billion annually to Australia's economy and support around 150,000 jobs by the end of the decade. The task is not to resist AI but to ensure Australians benefit from it.

That begins with skills. Supporting workers through technological change requires sustained investment in training and education. Through the National Skills Agreement, expanded vocational education and fee-free TAFE, we are increasing access to the skills Australians need to adapt and succeed. We're also investing in advanced capability in artificial intelligence and emerging technology.

The motion suggests that workers will be left behind. That ignores both the current policy settings and the role of our industrial relations system. Workers and their representatives must have a voice in how AI is introduced into workplaces. That is why the government is engaging with unions, industry and stakeholders to ensure deployment is transparent, safe and fair. Artificial intelligence can increase productivity and improve safety, but it also raises legitimate concerns around surveillance, bias and the nature of work. These issues are being addressed through consultation, regulation and the existing workplace protections.

Productivity gains should be shared, not concentrated. On regulation, the motion proceeds on an incomplete understanding. Australia already has robust, technology-neutral laws covering workplace safety, consumer protection and privacy. These frameworks continue to apply. At the same time, new risks emerge. That is why the government is establishing an AI safety institute to support regulators and ensure our legal settings remain fit for purpose. This is a measured and evidence based approach.

The government is also putting in place the conditions for investment and growth. Australia has a skilled workforce, a strong research base and access to clean energy. These are significant advantages in attracting investment in AI and digital infrastructure. The National AI Plan provides the policy certainty needed to support that investment and build domestic capability. The motion speaks of uncertainty, but, in reality, uncertainty arises when there is no plan. There is a plan. It is coordinated and practical and is being implemented in consultation with industry, unions and the community.

Ultimately, this debate is not about whether artificial intelligence will change our economy; it will. The question is whether the change will be managed in a way that reflects Australian values: fairness, inclusion and opportunity. This government's approach is grounded in those values. We support innovation, but we do not abandon workers. We invest in skills, not slogans, and we ensure the benefits of technological progress are realised here and shared by Australians. For those reasons, the assertions advanced in this motion cannot be supported.

Leon Rebello (McPherson, Liberal National Party) Share this | Link to this | Hansard source

When I started as the federal member for McPherson, in my first speech I made reference to the fact that government needs to have a vision that looks to the year 2100. We need to have a vision that looks beyond each electoral cycle, considers what Australia will look like in the future and, in doing so, considers the needs of Australians in the future and how we're evolving as a society. I think about my electorate when I look at this motion because I have Burleigh Heads in my electorate, which I'm very, very proud of, and—

An opposition member: A lovely place!

Burleigh Heads, in addition to being a lovely place, is an absolutely booming area for tech and for the tech startup industry in particular.

I recently had the opportunity to host Senator Andrew Bragg in my electorate, and we had a roundtable with a lot of the leaders of tech groups and new startup companies in my electorate. One of the things that came out of that really good discussion was that people felt government was behind the ball when it came to emerging technologies, including things such as AI.

AI is becoming a lot more familiar to Australians, and we're seeing it in our daily lives. But there is a part of it that also raises certain concerns. I hear from people, especially across small businesses, who are starting to see the implications of AI. Sometimes they're good and they enable us to be more efficient and more productive, but sometimes AI can also provide a level of concern in terms of job security. That's why one of the things that this motion touches on is the fact that the government has not developed an AI transition plan for Australian workers. I would have thought that, for a government that typically talks itself up about what it does for the Australian worker, it would have had some experience in putting something like that together—partly because we want to make sure that the government's got a strategy but also partly to provide a level of assurance to small-business owners in my electorate and across the country as well.

The scale of AI change is quite rapid, and we are starting to see routine tasks being automated. I think that as a parliament we should welcome that. We want to make sure that we're upskilling a new generation. I go into a lot of schools and universities in McPherson and see firsthand that the next generation is being upskilled in emerging technologies such as AI. We can use it productively, but we just need to get that balance right, and that's ultimately where I think this conversation needs to go. It's all about balance and about getting the balance right between embracing new technologies as a country, which we haven't always been very good at, and making sure that we are taking the Australian community and the business community with us on the journey.

The only way that Australians will benefit from AI is if this transition is managed effectively. The fact that there is no current plan for workers that would be displaced by AI is problematic, and I think it's incumbent on the government and the relevant minister to come into this place and tell us what their plans are, what their strategies are, for those individuals. We need to have a complex approach to that because we need to be targeting education. We need to make sure that those people are educated in alternative streams of employment. We need to make sure that they're able to use AI.

But we also need to make sure that new industries have the support they require, and we all know, especially those of us on this side of the chamber, that, when it comes to making sure that those industries are set up and act effectively, it's not typically government that has the best track record of doing so. It is the private sector that does that. So government should, in coming up with these strategies, work with the private sector. They should take lessons from, for example, the people that Senator Bragg and I met with—people who are at the forefront of this technology revolution. They're the ones government needs to be listening to.

Instead we've got a government that doesn't really know what it's doing when it comes to AI. They've had different messages coming from different ministers. We've had one saying we need to have a light-touch approach. Another one said, 'There are no plans to change it.' Then there's another one who has publicly called for a dedicated AI act. I think the people of my electorate and the people in the tech space across this country deserve better. They deserve a government that is prepared to lead so that we as a country can make the most of this transition for the benefit of Australians.

Claire Clutterham (Sturt, Australian Labor Party) Share this | Link to this | Hansard source

To put the concerns of the member for McPherson at rest, the government does have a national AI plan. There are three pillars. The second is to spread the benefits, and that has a dedicated goal of ensuring that Australian workers are not left behind in the AI transition. It also has a secondary goal of ensuring that Australian workers are appropriately upskilled and trained so they can meaningfully participate in the AI transition and share in the benefits. So the suggestion that the government doesn't have a plan for workers during the AI transition is incorrect. In December 2025, the National AI Plan was launched, and, as I said, it includes a direct goal of ensuring that Australian workers are upskilled and can share in the benefits that we will all enjoy through the AI transition.

AI is here to stay, and it has boomed in recent years. It has become commonplace in both business and everyday life. People use AI every day to make their lives easier, interacting with AI-powered virtual assistants or programs. Companies use AI to streamline their production processes, make unbiased projections of gains and losses and predict when maintenance will have to occur. AI can also help the Australian worker by reducing the risk of human error, and it can complete particularly difficult or dangerous tasks, meaning humans can mitigate or avoid the risk of injury or harm. So there are many benefits.

But, like anything, of course there are negative aspects. Although AI can create original and unexpected ideas, it just can't overtake humans in the ability to be creative, which means that it may be prohibited in its decision-making. If companies or governments or individuals are looking for a new or creative solution to a problem, humans are better at providing that solution, and humans, being human, will always inherently consider the emotional ramifications of decision-making. So there will always be that gap with AI, and there will always be a role for the Australian worker.

There is also a role for government and regulation in AI. It's the role of government to regulate without paralysing innovation and technological development. The role of government is to avoid overregulation because, as well as paralysing development, overregulation can create unnecessary fear and uncertainty. Governments have a difficult job of regulating to ensure that AI, which is not going anywhere, improves the lives of Australians in a way that is safe, sustainable and cost effective.

There is no greater need for balanced regulation than in the industrial and workplace space. This government will always protect jobs and working families—always. Our track record speaks for itself, with a consistent policy platform of protecting conditions for working Australians, growing wages sustainably, budgeting for continued rounds of tax cuts for every taxpayer and deliberately creating a pathway for collaboration and partnership between industry, unions and workers. The Labor Party will always stand with workers, and we believe that, by working together with unions, businesses and skills organisations, the interests of workers can be advanced.

Our consistent approach to conditions and support for working Australians will not change in the face of the disruption that AI poses in the workplace. We will support working Australians across a range of industries through all facets of AI—the growth, the disruption, the challenges. And we will do this by ensuring that worker protections remain a core policy driver, meaning that we'll consistently review and measure them to make sure AI is working for businesses and workers. We will also ensure that productivity gains from AI are shared fairly and not just concentrated. But, in doing that sharing, we will insist that the gains are real and measurable. We'll also invest in lifelong skills and training programs.

As well as being the party of working Australians, Labor is the party of education. We firmly believe that every Australian should have the opportunity to develop the skills and receive the training that they want and that we need in order to prosecute our policy of a Future Made in Australia and so that working Australians can meaningfully participate in a changing workforce in order to earn a fair wage, contribute to the economy and provide for their families. This is why we have a National AI Plan, which has a deliberate AI vision for Australia, where AI is used to build a fairer, stronger Australia and where every person, especially working Australians, benefits from this technological progress.

Garth Hamilton (Groom, Liberal National Party, Shadow Assistant Minister for Energy Security and Affordability) Share this | Link to this | Hansard source

It's a very important topic that we're speaking on—artificial intelligence—and I'd take a moment to acknowledge the member for Chifley, sitting opposite, and the contributions he's made to this and to my local manufacturing industry, which relies heavily upon AI and will continue to do so. So when I speak to this topic I'm speaking about local jobs. I think there are many people across the chamber who still focus on good jobs and good pay, and that's what should be driving this current debate.

We have seen the largest drop ever in productivity in Australia since Labor came to government—over a five per cent drop. We could have an argument as to the causes of that. The Labor government will point to events that have happened outside this nation, we will speak to policy decisions that the government has made, and we can have that debate about the causes of that. The important point, though, is: what are we going to do about it? What is the government of the day going to do to address that incredible drop in productivity—over five per cent?

We're blessed, in some ways, that at this time in Australia's history the opportunity AI presents has made itself available. We have to acknowledge that this is a race across the world. The people who want to build data centres that would support localised AI within each nation have constraints on their capital and are having to make decisions based upon where they can invest and who's going to support that the most. The countries that are able to best support the investment in data centres are going to have the most benefit. It's a very simple equation. And, given that drop in productivity, it's absolutely vital that Australia jumps into that race and leads it.

I was recently speaking to the headmaster of Toowoomba Grammar School about the role AI plays in the education system there. They've gone from a time when they would actively ban the use of AI in any assessments to considering how to include the use of AI in the assessment of kids—knowing full well that the jobs of the future will be for people who can use AI, who can use it well and who can understand it in the same way those of my generation learnt how to use the internet and those of previous generations learnt how to use AutoCAD or whatever was the driving technological advantage of the day. AI is going to be a tool that we have to use, and we have to get into that race. It's as simple as that.

I want to talk to what I think is the most important part of that conversation, which is cheap energy. Regardless of the moral views that come out in the energy conversation, if we want AI data centres to be built in Australia then we're going to need to provide them with cheap energy, or they're going to have to provide it for themselves. That's the reality that is faced all over the world, and everyone we speak to who is interested in investing in data centres has a No. 1 concern: what is the cost of energy? This is not a technology that can rely on intermittent power. This is not a technology that can just run when the sun is shining or when the wind is blowing; this is technology that needs to run 24/7. If we don't do that, we will not have the advantage of it. Of course, there is the question that's obviously put forward: what about batteries? To which I would answer: South Australia's battery is charged by coal and gas from New South Wales and Victoria. We see those numbers from AEMO every single day.

The cheap energy that Australia needs now and will need to get in front of the race for the development of AI data centres will come from things like coal and gas—and, goodness me, who knows, if we can get over ourselves and look around the rest of the world, maybe even nuclear. This is how the race is being won by other countries right now. We have a very short window to decide whether we actually want to be in that race or if we want to sit back and be bystanders and watch others take advantage. The only thing that we can offer at the moment is the stability of government. That's a good thing; of course, it's Australia's great advantage.

We are not building the structural advantages into our system of government as to why we would build an AI data centre in Australia. We do not provide cheap energy. The number one thing that will be needed, we don't provide. Until we do and until we make the firm decision that that's important to us and that we don't want to miss out on the next great leap forward that is happening around the world, we will fall behind; we will not win this race.

Ed Husic (Chifley, Australian Labor Party) Share this | Link to this | Hansard source

In the last few months, we've seen thousands upon thousands of tech workers retrenched due to AI—workers who were right to believe and who had been made to believe that they'd have a job for life. AI has busted open the sanctuary gates. If those jobs are no longer sacred in an AI future, then what about the rest of our workforce? Because, while a serious danger before us is that a lot of jobs will profoundly change or disappear due to AI, the greatest danger is that, as Professor Toby Walsh has put it, we will sleepwalk into an AI future not shaped by us at all. It's a question of power, of control, of sovereignty.

Governments have a responsibility to think ahead, step up and act in the national and local interest, looking out for communities that might be impacted. Governments can't just be cheerleaders for novel uses of AI; they must prepare and dilute the associated risks. We can't be scared to act because we're worried about the reaction of the Trump administration. We can't have a hands-off, laissez-faire approach to AI or respond in a knee-jerk, spasmodic way to the AI risks that only threaten the loudest or most powerful voices in society. Saying we want overseas generated AI models to bake in Australian values is a terrific rhetorical sweetener, but, without any legal or enforcement power, it simply melts in the heat of big tech obstinacy. We need a comprehensive, national, economy-wide AI act that identifies risks and sets out our expectations on how to manage them and build sorely needed confidence in AI.

Sixty-five per cent of Australians believe that, overall, AI creates more problems than it solves. That's up eight percentage points since 2023, according to Roy Morgan Research. Linked to this mad rush around AI is a frenzy to build data centres. There are now more than 250 data centres operating in Australia. We attracted $10 billion in data centre investment in a single year. But this expansion is not without consequence. Data centres are extraordinarily energy intensive. Their impact on carbon emissions doubled over five years. They could represent 11 per cent of Australia's electricity consumption by 2035, up from one per cent today.

The demand on our labour force to build this new wave of data centres is also significant at a time we're 90,000 workers short in construction and trying to build 1.2 million homes by 2030 and roll out more renewables and deliver a major infrastructure pipeline. Many of these overseas tech firms want governments to move faster, water down regulations, speed up access to land energy infrastructure, while, in some cases, even threatening not to build data centres if they're forced to pay tax. Those points bring into scope fundamental questions about the social licence of many of these tech firms.

I want to be clear: I am not anti data centre, but I am anti hype, and I'm very pro getting our act together to devise a structured approach to infrastructure, energy, workforce, taxation and sovereign capability. If overseas tech firms want to build data centres here, then spell out the benefit for the country. Will their computing power be set aside for Australian industry and researchers to help Australia tackle our challenges? That should be a condition of the Foreign Investment Review Board when they attach approvals on data centres that overseas firms want to build here. Overseas tech should be required to show that, when they build data centres, they're not going to negatively impact on the power and water supplies of our communities. They should demonstrate they won't compete for labour, especially in regional Australia, which is helping build more homes or infrastructure work that's vital to the nation.

There's something else that needs to be taken up. Many data-centre builders say they'll build new renewable supply to meet their needs. They can't always do that onsite. They have to draw supply from elsewhere. That means energy needs to be moved around. Data-centre builders should be required to pay the cost of that, specifically via transmission levies, contributing to the cost of building transmission lines. That's not radical. Canadian provinces have introduced a data-centre levy applied to data centres connected to the grid with electricity capacity of 75 megawatts or more. This means power-hungry facilities pick up a slab of the cost associated with the energy demand that they created, shielding residential and business customers from paying for this through their energy bills.

We need to debate AI openly and have a wider public discussion about artificial intelligence and the infrastructure needed to support it. Yes, we have to act fast. Yes, AI has benefits, but we can't allow AI to reshape our economy or sovereignty on terms the Australian public does not choose or support.

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

At its heart, this motion speaks to the government's current hands-off approach to artificial intelligence. AI is already reshaping economies, labour markets and the information environment, and, while its ultimate trajectory remains uncertain, the scale of potential change is enormous. Yet, despite this, Australia has done remarkably little strategic thinking about what this transformation means for our economy, our institutions and our society. That gap was apparent in the government's long-awaited national AI plan. It set out three worthy objectives: capturing the opportunities, spreading the benefits and keeping Australians safe. But, beyond that, its 37 pages largely collated existing announcements. Somehow both the techno-optimists and the techno-pessimists walked away dissatisfied. There was little ambition in relation to capturing AI's upside and limited reassurance on safety.

While I commend the establishment of the AI Safety Institute, the plan did not offer much of a sense of preparedness for the future. The underlying message felt like, 'Let's sit back and see how this develops.' That approach is also reflected in investment. Australia is not investing much in AI compared to other countries. Over the last five years, Canada has invested six times more than us in AI. Singapore has invested 15 times more than us. The UK and Germany have both invested three times more than us per capita. To be fair, part of this tentative approach reflects genuine uncertainty. No-one knows exactly how powerful AI will become or how quickly. Perhaps today's excitement will amount to little more than a generation of very effective chatbots. But, globally, there's an incredible amount of capital riding on the bet that the impact of AI will be much greater than that.

We should be shaping our own future. At present, many of the most consequential decisions are being made by the tech oligarchs on the other side of the world. We must seize the reins. So how do we do this? First, we must deal with the risks and opportunities that already exist. We're already seeing real harms from AI: psychological distress from chatbot interactions, sophisticated AI-enabled scams and fraud at scale. A sensible starting point is a digital duty of care, requiring platforms to take reasonable steps to prevent foreseeable harm. This must extend to include AI chatbots. We should ensure our regulators have the technical expertise to identify and disrupt AI powered scams. We must also start to develop policy to manage AI enabled disinformation campaigns, particularly during elections.

On the opportunity side, AI's potential in science and research is already clear. Australia should be investing heavily here. The government has already committed more than $360 million through existing programs. That's a strong foundation, but it must be clearly directed towards AI powered research, and it must be scaled up.

Second, we should pursue no-regrets policies, policies about our future that make sense regardless of how AI ultimately develops. Strengthening the AI Safety Institute is an obvious one. The institute has a significant responsibility in monitoring risks from AI and working with policymakers and regulators to manage them. Yet funding for our AI Safety Institute is about one-sixteenth of comparable efforts in the UK. It needs more funding. It also needs to be protected from the bureaucracy of the standard Public Service so it can move nimbly and independently. The government should also consider giving the institute more powers to gather information so it can effectively monitor the AI risk landscape.

Another no-regrets policy is investing in an AI-ready workforce, from university training to mid-career transition pathways and deepening international collaboration with trusted partners on standard safety and governance. These investments will never be wasted.

Finally, we must start planning for the most significant future risks. There are some very dramatic predictions about how AI could change our workforce, society and economy. These may or may not come to pass but, with the speed of AI, we can't afford to wait for them to happen before thinking about how we might respond. Even if the likelihood of some of these scenarios is low, the consequences are so significant that we need to be prepared. A large chunk of the workforce could become unemployable. Economic value could be concentrated in a small number of companies across a wide range of industries, and our tax bases could be undermined. All of these shocks would require a government intervention, and it would take time to build the social licence needed. What is clear is that this hands-off approach is not good enough. The government must take an active approach, because we must ensure that Australian voices are determining our future, not big tech.

Cassandra Fernando (Holt, Australian Labor Party) Share this | Link to this | Hansard source

The time allotted for this debate has expired. The debate is adjourned and the resumption of the debate will be made an order of the day for the next sitting.