"Teens are the next generation to inherit the decisions we make today, yet are rarely consulted in those decisions. My hope is that alongside consulting industry experts, the Office of AI meaningfully includes young people and consults their expertise.

Young people are among the fastest adopters of AI and are navigating an experimental technology in real time, with a mix of curiosity and anxiety. They express worries not only about their own use, but about how adults use, design and govern AI. They are looking to us now. We need their input not only to ensure we’re centring their needs and safety, but also so we can learn from their creative and innovative ideas."

Annabel also provided the below quote from an anonymous teen in the AI Youth Council:

"AI should be made by governments, not by companies. The AI world seems highly uncertain, with the leaders for AI making and developing AI for the purpose of money and pleasing shareholders rather than the people that actually use it."

"The Australian Government addressing AI is to be welcomed. The opportunities presented by AI are, as with any gold rush, mostly in providing shovels. That is, the economic opportunities are not just in providing the AI itself, but in ancillary services around it, and in implementing it in practical applications. Australia is a leader in international education, mining and agriculture, all areas where AI can be applied. Australia can use AI to improve the efficiency of its own education, mining and agriculture industries, and also export that expertise to the world."

"AI is a transformative technology that is being adopted worldwide across business, government, healthcare, finance, critical infrastructure, and academia. Australia’s growing reliance on non-local, closed-source commercial AI services presents risks to data security, intellectual property, and long-term global competitiveness. The AI policy described by the Prime Minister comes at a critical time, laying the foundations for Australia's sovereign AI future and outlining the regulatory framework for the sustainable construction and operation of data centres. The new Office of AI should prioritise the development of sovereign AI capabilities alongside effective regulation of data centres."

"The establishment of a centralised Office of AI alongside a single, national framework is a vital step toward bringing cohesive governance to Australia's rapidly evolving tech landscape. Historically, fragmented regulatory responses have left local industries balancing immense opportunity with heavy uncertainty. By organising our rollout through the Department of Prime Minister and Cabinet, the Government is signalling that AI is now treated as critical cognitive infrastructure rather than just an isolated technical tool.

Crucially, the decision to firmly deny copyright waivers to AI firms mining local data is a massive victory for Australian sovereignty and knowledge protection. Tech giants have long operated on a 'mine first, ask forgiveness later' model, often exploiting local content to train models with impunity. Drawing this line in the sand forces developers to respect intellectual property and respect ethical data boundaries. However, the true test of this announcement will lie in its implementation. For the national framework to succeed, the new Office must bridge the gap between policy and practice—ensuring we actively fast-track data infrastructure and incentivise ethical local innovation, while remaining agile enough to protect public safety as machine learning capabilities rapidly advance."

"The move announced today about establishing Office of AI is a step in the positive direction.

AI is a critical technology that governments need to take seriously as it will have enormous impact on all aspects of our society.

However, I must say that the announcement is about three years late. Other jurisdictions have already made significant progress around regulating AI, most notably the EU with its AI act from August 2024.

While Australia is lagging behind in this domain, I’m hopeful that this would signal that AI will be taken seriously by the government and lead to much better engagement with the scientific community to further advance the state of AI in Australia.

This also means more funding for basic AI research, and also applied domains such as AI in Education, Health, and other areas."

"I welcome the initiative to streamline, integrate and improve the efficiency of data centre approvals. One of the largest constraints on AI innovation is access to compute, including data centres to power AI systems. As such, the focus of many global AI investors is developing infrastructure.

Due to its strong rule of law and stable economic environment, Australia has a strong competitive advantage. Technology companies do not merely seek tax incentives, they seek regulatory certainty. Clear, predictable approval pathways and governance structures that reduce investment risk.

Australia has an opportunity to distinguish itself by offering not only fast approvals, but trusted digital infrastructure, supported by trusted legal protections around data privacy, cybersecurity and AI governance.

But - we need to balance this support of AI investment with measures to protect our own national interests and maintain our digital sovereignty.

For example, we need to ensure data centres are included as critical infrastructure within our security of critical infrastructure laws, and ensure we have in place mandatory risk management programs and reporting obligations.

We also need to ensure that data centres uphold Australian cybersecurity and data privacy laws. This is especially important in the context of cross border data flows and offshore data storage. We need to ensure that there are no cybersecurity vulnerabilities created by hosting data centres in Australia.

So, we need to balance the support of data centre development with addressing relevant privacy, cybersecurity and AI governance issues, as well as environmental concerns.

While there are intellectual property issues relating to the unauthorised use of copyright works to train AI systems, we are still awaiting the outcomes of multiple major global lawsuits. It is not an area that Australia can effectively resolve in isolation.

If Australia can get the balance right, it will become an attractive destination for AI investment that will help drive productivity and economic growth."

"There are several significant gaps in the PM’s speech on AI planning that warrant public attention.

On Australia’s digital sovereignty and the role of big tech: the speech reads more as a statement of reliance on the promises of multinational technology firms than a set of enforceable measures. The question of whether Australia can trust companies like Amazon, Microsoft, and Meta building data centres here or using their models which have already been embedded in business and public services, remains unaddressed. Digital sovereignty is not simply about where data is stored; it is also about ensuring that data generated in Australia does not leave the country’s jurisdiction (see China’s cybersecurity laws). The speech offers no concrete framework for this.

On employment and AI economy: while the PM cites graduate employment rates as a positive indicator, the data alone says little about the quality of jobs themselves and the connection of current employment in relation to the AI economy. This is not a new problem. Predictably, the quality of jobs and workers’ experiences are declining, drawing upon current research and past research in the digital economy. The platform economy has already demonstrated how digital transformation can degrade working conditions, extending hours, entrenching precarity, and subordinating workers to algorithmic management and surveillance with the societal consequence of widening class stratification. Australia was slow to regulate gig work and existing research shows that the AI economy is compounding these trends. Workers in the data annotation industry who label training data for AI models to learn are a case in point. Once again, protective regulation is not yet deployed in anticipation of the harms already taking place."

"The establishment of the Office of AI recognises that the Federal Government has an essential orchestration role to play, moving us beyond what has until now been a fragmented landscape across different levels of government, regulators and sectors.

As Australia develops a national AI framework, it will be crucial that policy is underpinned by a strong, independent evidence base, bringing together government, industry, universities and civil society.

This was a commitment that I was encouraged to see reflected in the PM’s announcement today.

Building that evidence capability is going to be just as important as the governance structures themselves.

The university sector has a crucial role to play in providing independent research, impact assessment and trusted expertise to help ensure AI delivers long-term economic and societal benefits for Australia."

"The speech was strong. What I most liked was that it kept Australian values, sovereignty and safety at the centre.

Establishing an Office of AI within the PM’s Department is the right move. AI governance is currently scattered across regulators, institutes, plans and voluntary frameworks. Someone needs to bring those pieces together and take responsibility for the whole picture.

The new Office will need four things: speed, genuine expertise, teeth, and adaptability. Australia is already behind, and the technology is moving faster than the government. We do not have time for months of political point-scoring. Australia now has a new wave of PhD graduates whose research has focused directly on generative AI. They are genuine experts in its risks, capabilities, evaluation and governance, and they need to be inside the strategy process from the beginning. We also need enforceable requirements, clear responsibility and consequences when systems fail.

Just as importantly, policy cannot be built only for today’s AI. Most current generative AI still descends from the transformer architecture introduced in 2017. Agentic AI is already a step change. These systems can plan, remember, use tools and act over time. If we are only now preparing to govern standard generative AI, we are nowhere near ready for agentic systems.

Australia may never build the world’s biggest model. But we can lead on the harder problem: how powerful and agentic AI is evaluated, governed and made answerable to the public in ways that reflect Australian values."

"An Office for AI is past due. Australia needs to stop playing catch-up and develop a national strategy to reduce our reliance on foreign technology and infrastructure. As the exponential increase in AI capabilities disrupts our personal and professional lives for better and worse, we are sacrificing our privacy and sovereignty at the expense of short-term productivity gains.

We need a concerted strategy for the safe and secure adoption and development of AI technologies in line with our national interests, ensuring that Australian resources and assets are not exploited at our expense. The potential for AI to enhance our lives is enormous—we must ensure this is done with trust, respect and transparency."

"Australia's Office of AI represents a game changing opportunity for Australia to set AI guidelines, policy and legislation that will affect academia, industry, government and all of society.

One of the most crucial areas where AI regulation is needed is in the production of potential military research or weapons of mass destruction.

Existing laws and regulation in Australia in the AI space are simply not up for the challenge of regulating the rapid pace of technological development.

The best example is using generative AI to develop a biological weapon - something that might not be illegal under Australian law.

The analogy would be like ordering all of the parts of a machine gun and leaving it on your coffee table, unassembled, and somehow that is not a crime."

"The establishment of an Office of AI is a positive step. Australia should not approach AI only as a risk; it is also a major opportunity in education, business and healthcare. In remote communities, an agentic tutor could provide personalised explanations, practice and feedback where specialist teachers are scarce, while supporting—not replacing—the teacher. In healthcare, AI can help clinicians detect patterns in scans, support earlier diagnosis and reduce pressure on overstretched services.

However, access must be matched by responsible use. In education, excessive dependence can lead to cognitive offloading, where students hand over the reasoning they need to practise. This can create an illusion of competence: the work looks impressive, and the student feels capable, but may be unable to explain or reproduce it without AI. Over time, this risks weakening critical thinking and metacognitive skills—the ability to monitor and correct one’s own reasoning.

That becomes serious when future doctors, lawyers or other professionals rely on AI without verification. Australian legal proceedings have already exposed the consequences of unverified AI use, including submissions containing fictitious cases and quotations. A national framework should therefore promote AI adoption while requiring human oversight, verification and safeguards in high-stakes settings. The goal is not less AI, but AI that strengthens human capability rather than replacing it."

"The establishment of an Office of AI within the PM Department is a significant step as it moves Australia towards a coordinated national approach to AI governance.

In my view, if it is implemented effectively, it has the potential to provide greater certainty for businesses, researchers and the public. The office may need to adopt a risk-based regulatory approach that applies stronger oversight to high-impact AI applications while allowing lower-risk innovation to flourish. The government's decision around copyright for AI training will also send an important signal that protecting creators' rights should remain a core principle, while encouraging the development of practical licensing and data-sharing mechanisms that support responsible AI innovation.

At the same time, governance alone is not enough and Australia must continue investing in AI skills, research infrastructure, sovereign AI capability, and stronger collaboration between government, universities and industry to ensure the country is not only regulating AI effectively but also building the capability to become a global leader in responsible AI adoption. I believe this office should be able to deliver this."

"Prime Minister Anthony Albanese’s announcement that "Australia is the first country worldwide to bring economic, social, security and environmental issues of AI under one office" is highly ambitious and appears to be missing a critical detail. Each of these four issues imposes challenges and risks that conflict with each other. For example, the deep public distrust in AI software/adoption (in job losses and deepfakes), mistrust in AI hardware (in data centres, IP of training data), lies in conflict with the generational economic and wealth creation opportunity for capital investment in data centres, renewables, and energy infrastructure.

We could speculate this announcement is driven largely by the economic opportunity presented by frontier AI labs wanting to build their own data centres here in Australia. As Anthropic has indicated, such an investment is contingent on the text and data mining exemption from our copyright laws that allow these models to be trained using public data sources, regardless of copyright status.

The Prime Minister’s reference to "Australia becoming a ‘data warehouse’ for overseas models" is further confirmation of this contingent factor. We should make no mistake that the best AI labs in the world are big American businesses with full US government support, and they will dictate many of the terms locally.

By acknowledging these conflicts and challenges, this new Office for AI and the government could find the right balance not to miss out on the economic and productivity opportunity of AI, while also ensuring social licence and community support for this sector.

We should learn from international experiences: Ireland for poorly formed data centre strategy; New York, an industrial city with a very recent data centre moratorium; San Francisco for deep AI investment; and the European Union for AI regulation.

We're not the first in this space, but we have clear pathways to find the right balance. Start with world- leading AI literacy for all Australians, as most of the voting public have very limited appreciation of the potential of AI. And Universities have an important role to play in this."

"The Prime Minister is right that Australia should be more than a data warehouse for AI made overseas. But you don't stop being a data warehouse by regulating data centres. Medicare wasn't a standards framework; we built it. The Snowy Mountain scheme wasn't a policy announcement; we built it. If we want Australian AI, we have to fund the people who build it: our universities and our companies.

You cannot regulate your way to a sovereign capability. Setting the terms of someone else's technology isn't sovereignty, it's tenancy."

"The establishment of an Office of AI and a commitment to a single national framework are both sensible steps. AI governance benefits from national coordination rather than a patchwork of state-based approaches.

The speech was obvious in sending a very deliberate signal to global industry. Australia was presented as a politically stable democracy with reliable institutions, abundant resources and a comparatively straightforward regulatory environment. At times it felt less like a policy announcement and more like an investment pitch: Australia is open for your AI investment. This raises important questions about what Australia expects in return, particularly around energy use, environmental impacts, water consumption and benefits to local communities.

The government acknowledged too, though, that AI infrastructure has real environmental consequences. Too often we talk about AI as if it exists purely in the air, when in reality it's built on highly resource-intensive physical infrastructure. There was some signalling around neutral impacts on energy grids, however we still need clear expectations around renewable energy commitments, environmental accountability and ensuring these developments deliver a genuine public benefit rather than simply private returns. While we heard much about Australia's abundance of sunlight, there was little detail of specific renewable or net-positive environmental targets, this detail will be crucial.

The government's position on copyright is also significant. Rejecting a blanket copyright exemption for AI developers sends an important signal that Australian creators shouldn't be used as an unpaid training resource for large technology companies. But copyright has been under strain for years, long before this current generation of AI technologies arrived. If we genuinely want a thriving creative sector, protecting existing copyright law is the lowest bar to clear, we also need broader cultural policies and investment that improves the economic sustainability of creative work. On this, there was nothing in today's announcement in how we properly invest into and create a thriving creative sector in the AI age.

One last aspect of the speech that stood out was its strong emphasis on AI sovereignty and the idea that Australia needs to build AI domestically in order to shape its future. That's an understandable aspiration, but it risks becoming technologically deterministic. Australia has always been able to shape markets through regulation. We don't manufacture every product sold here, yet companies routinely adapt their products and services to Australian laws and standards. The same principle applies to AI. Building domestic capability is valuable, but our capacity to govern AI shouldn't be seen as contingent on owning every part of the technology stack."

"It is a great opportunity to develop renewable energy and update the classic fossil fuel energy system in Australia, as well as the development of Australia's abundant resources. Data centres for AI need a very stable electricity supply. Renewable electricity is preferred, but it must be backed up with the stable electricity grid from current fossil fuel power plants.

Australia is not only rich in renewable energy, it's also among the world's richest in fossil fuels like coal and natural gas. Coupling with carbon capture technology, the huge financial investment from data centres will update our current power plants to carbon zero energy generators. Australia's grid is outdated and limited by the current manufacture development. It can be updated by the investment from AI. In addition, the new data centres release huge heat during the process, so new heat transfer materials based on Australian mining resources are required. It will help the mining industry and manufacturing in Australia."

"This is a good call by the government, although it should have happened earlier. Having a national framework for the technology is a must to maintain Australian sovereignty, in the cyber and, consequently, the physical world. AI agents operating in Australia and for Australians need to be certified by Australia, and the national framework is the first step toward this. I am very supportive of this move."

"Australia is right to want a seat at the table as AI reshapes the global economy. Australia is right to make data centres our ticket in. Few nations are better placed than Australia: abundant land, abundant energy, political stability, and trusted security relationships. Fast-tracking data centre approvals plays directly to those strengths. For us, countries that host AI compute get a seat at the table when negotiating the future.

Copyright has been the sticking point. Our laws were never designed for AI training, so models get trained overseas and Australian creators receive nothing. If we don’t sort out copyright so models can be trained here, our creatives are no better off. They still lose their IP but also lose their bargaining power to do anything about it.

Recent national polling found 61 per cent of Australians want copyright arrangements that enable AI training here while still supporting creators; only 15 per cent want the status quo. A scheme where AI companies training in Australia pay into a fund for creators would beat the current worst-of-both-worlds. I am encouraged the Government looks ready to strike that balance."

"I often think of AI as a double-edged sword. It is an incredible opportunity to solve many extremely important problems across a wide range of fields – there is no doubt that without machine learning in particular, we’d be worse off, with fewer opportunities to develop vaccines, cancer drugs that work and so many applications to assist in our adaptation to climate change (for example - IBM and NASA collaboration that has given open-source models to scientists for climate risk modelling work).

There are upsides and downsides – possibilities that have great outcomes and unintended consequences. In the past, Australia has squandered our lead by not recognising and supporting the brilliant and innovative scientists that work in our Universities on difficult problems where the outcomes are not easy to predict.

The Prime Minister's announcement today brings hope for a future where we recognise the risks and build for the kind of Australia we want - where AI supports our democratic processes, ensures equity in access to rewards and protects our human rights. But to realise this, the diversity of voices we listen to needs to be prioritised. I hope this is realised."

"The prime minister’s announcement is welcome. Targeted action on harnessing the AI opportunity is exactly what we need. Streamlining the activities through a single framework, protection of the creative industries and turbo charging renewables through data centre investment are all welcome principles. Keen to see more details on how that would all be executed.

Australia’s investment in AI has been trailing behind peer countries and uncertainty in the regulatory landscape makes businesses uncertain. We hope that the new AI Office will mobilise the much-needed investment into local AI capability development. There is immense talent in the country- deep tech expertise in the universities, and bold innovative ideas from the startup community - we hope that the AI office will call upon representatives from these groups to provide input and advice on shaping the future of the AI ecosystem in the country."

"Generative AI absolutely needs regulation and monitoring, but given how the technology works, it's impossible to make safe.

Calls for 'safety rails' and 'ethical use' aren't possible.

I applaud the protection of Australian works, but this is only a start. We already know that generative AI is producing dangerous misinformation at an alarming rate – in particular this targets trans, gender-diverse, and queer people."

"It's incredibly positive for the Prime Minister to be talking about respecting the copyright of Australian artists, and ensuring that Australian values are reflected in AI. However the devil will be in the details, as Australia is well behind other countries when it comes to investing in Australian AI capacity.

And the investment in people will be crucial. For all the talk of policy solutions, there are real, significant technical and legal challenges required to ensure that AI companies serving Australians respect Australian copyright law. AI companies have a real incentive not to play ball with Australian copyright law, because the majority of these models can only exist thanks to copyrighted material. If Australia wants to force a fair go for artists and creators, we may have to partner with other nations to make this happen. We should be looking to the EU's AI act, and how it forces any AI companies serving European users to fully disclose where they sourced their data from.

One thing that really seemed missing from the speech was a focus on Australian knowledge. Data centres are huge capital investments, but between 60 and 90% of the money spent in them goes to overseas companies. We don't make the chips, the cooling systems, and the hardware that makes AI data centres a reality. If Australia really wants to capture the possibilities of the AI era, we need to be investing in our people - building up capacity in research, in AI start ups, and the broader workforce to ensure that we are a place where the value of AI is unlocked, rather than just where it is stored."

"Today, Albanese wrote the mathematical equation we all need to remember: Artificial Intelligence = Australia’s Interest. The Prime Minister presented a very welcomed strategic move. Australia can grow its strategic role in the AI space by accelerating investments in core foundational capabilities: data centres, workforce, and inflow of investors. Together, they support the four resources necessary for any strategic capability: Land, Capital, Labour, and Knowledge.

As an AI researcher, I spent my life on the science. Without doubt, data centres form the foundational capability required for today’s AI technologies that are data-driven, such as large language models. Risks should not be underestimated. It is so crucial to get the standards and regulations balanced. Not all AI technologies are data centres based. In addition to copyright, we need the standards and regulatory frameworks to cover also the whole supply chain where these data centres are just one hub. We should not export the coal to buy it back processed unless our decision is evidence-based; we need to have the processing capability locally (computing, algorithms, trained people, clarity on economic benefits, policies, and social license).

Data Centres are possibly the start of the journey; we need to ensure we have a good vision of how the journey may continue (strategy) and be adaptive to rebalance the equation in this vast-changing landscape. To match the speed AI is advancing internationally, we need a whole-of-nation effort. Today, Albanese told the nation, we are one Australia going for one AI strategy and standard journey. What will follow needs to be balanced from the lowest written document, to policies, processes and procedures, to efficient and streamlined operations, to investments, and to the social license where the Australian people ride the vehicles of this journey."

"ATSE has welcomed the Australian Government’s new Office of AI. It must focus on investing in the people, standards and infrastructure needed to compete globally. The new office is also an important step towards nationally coordinated AI policy but governance alone will not secure Australia's long-term competitiveness.

Legislated national AI standards reflect the recommendation in ATSE's AI Investment Blueprint, released last year. The Blueprint also calls for greater investment in Australia's AI workforce, sovereign capability and enabling infrastructure.

ATSE's Investment Blueprint found AI is now a source of national competitiveness, with global AI investment expected to approach AU$3 trillion this year. Australia has invested just over AU$300 million over the past five years, while several comparable economies have committed more than ten times that amount.

ATSE welcomes the Government's intention to establish nationally consistent requirements for data centres, including energy and water obligations. As demand for data centres grows, investment must also support renewable energy, water-efficient technologies and technologies that can make this digital infrastructure more sustainable and that deliver long-term benefits for local communities.

ATSE looks forward to working with the Australian Government to strengthen national coordination of AI policy and ensure Australia captures the economic opportunities presented by AI while maintaining public trust.

Once the regulatory framework is developed the next challenge is ensuring we have the engineers, researchers and skilled workforce to turn AI opportunities into Australian capability, Australian jobs and Australian exports."

Additional background information: ATSE recently made a submission to a NSW inquiry into data centres. ATSE has also published a Made in Australia AI action statement.

"Prime Minister Anthony Albanese’s speech was a masterclass in explaining to US Big Tech why Australians trust governments to keep them safe. He outlined Australia’s proud tradition of innovations in fairness – from universal healthcare in Medicare, universal secret voting, universal superannuation, the eight hour [working] day, minimum wage, and the Under-16s social media ban.  This has made our country one of the most stable and wealthy democracies in the world.

The AI companies have committed the biggest act of copyright theft in history. The PM firmly told them that not everything here is "up for grabs". He said that writers, journalists, and artists "must retain ownership" of our work, which is our property. Books, music and art he said, have been taken by Big Tech AI companies without artists’ consent and that the artists need control over the price and value of our work. "Anything less", the Prime Minister concluded, "is theft".

This is a clear message to Big Tech that if they want to operate in Australia, they need to obey our laws, including copyright law. They need to come to the table now to negotiate with writers and other copyright holders. Anything less, is theft.

The PM left open the possibility of a reparations fund for the past theft by the AI companies of our work."

"Standing this office up immediately sends a clear signal that AI is now a whole-of-government priority, not a side issue for one department to manage alone. That matters because capability alone rarely drives adoption of new technology. It stalls just as often because no one is clear on who is accountable when a system gets something wrong and a business has already acted on it. A coordinating office is a real step toward closing that gap, provided it’s built with the industries and workers it’s regulating rather than issuing standards to them from the centre.

The firm line on copyright is just as important. Refusing AI firms a waiver to mine Australian data protects the creative and knowledge industries most exposed to this uncertainty, and it tells businesses the rules protecting them will hold. That trust, more than speed of approval, is what will determine whether this framework actually works."

“Data centres are becoming essential national infrastructure. Australia needs local computing capacity to support research, industry, government services and data sovereignty rather than relying too heavily on overseas systems. But supporting data centre development is only the starting point.

The bigger question is whether Australia has the energy, water, workforce, grid capacity, cybersecurity protections and long-term planning needed to support this growth responsibly. By the time a major data centre is planned, approved and built, the technology, energy demands and business models may have changed significantly.

Australia therefore needs agile governance, with clear protections, regular reviews, measurable targets and the ability to update requirements as new evidence emerges. There are also important opportunities and trade-offs that are not yet fully understood.

Data centres could support innovation, jobs, research and regional development, but they could also place pressure on electricity grids, water supplies, local communities and emissions targets. The government must have a clear process for identifying these impacts before decisions become costly or difficult to reverse.

That means transparent assessments of energy, water, emissions, employment, privacy, copyright and community impacts, along with public reporting and genuine consultation with researchers, creative industries, workers, local communities and First Nations peoples.

Social licence will not come from announcing new infrastructure or making broad promises - that will depend on whether Australians can see who benefits, who carries the costs, and whether the government has the capacity to respond as the technology changes."

"I think the announcement’s focus on proposed energy and water requirements for new data centres is important. However, Australia should also encourage research into, and adoption of, genuinely efficient AI: meaning systems that are faster, less resource-intensive and more sustainable.

The main reason for that is AI systems are scaling rapidly and becoming increasingly computationally demanding. If there’s no standards and incentives for efficiency, the competitive benefits and low marginal cost of using AI could drive very high aggregate resource consumption. Infrastructure policy should actually look beyond the efficiency of data centres themselves and consider the models and software running inside them.

Also of course there’s a sovereignty problem. Foreign AI companies will increasingly compete for Australian computing, energy and water resources. If Australia does not develop efficient local AI capability, governments and businesses may become more dependent on foreign-controlled platforms for advanced models and computing capacity. From my research on efficient LLM and retrieval-augmented generation systems, I see efficiency as a design choice, not only a hardware issue. National standards should encourage transparent reporting of computation, memory, energy use, latency and performance quality for adopted/planned tasks. This would help Australian universities, startups, public agencies and smaller companies deploy capable AI with fewer resources."

"Overall, I am pleased to see the government committing to a national AI strategy. I think there are excellent points, such as the environmental regulation of data centres that are critical to prevent what has been happening in countries like Chile, for example. My concern is in the details of how the strategy will be executed. At the address, the PM himself said that the social media ban was about sending a signal. Now we are seeing that the problems with that policy were in how it was actually implemented.

We have a similar situation here, where the devil is in the details. Protecting artists' copyright is important, but it is technically very difficult to enforce or even remove that data from the model. He mentioned building sovereign capability, but Australia has historically underinvested in AI relative to other rich nations. He mentioned developing national standards, but we've been here before. Around 2018/2019, several organisations produced close to 100 sets of AI ethics principles, but these principles didn't have any teeth and were often in tension with each other.

For example, we all want AI to be transparent, but also private, but explaining an AI decision can reveal the data behind it. Human oversight of AI decisions sounds great, but does not scale. The challenge for the government is in creating a set of standards that are flexible enough to accommodate the evolution of this technology, while being specific enough to be enforceable"

"Today, the Prime Minister of Australia spoke at AI in Australia's Interest about the future of AI in Australia. The talk covered almost all the key issues surrounding AI, including safety, economic opportunities, and its impact on jobs.

The point that stood out to me most was the commitment to protecting the intellectual property of artists, journalists, writers, musicians, and other content creators. The Prime Minister stated that Australian creators must retain ownership and control of their work, and that no company should be able to use their content to train AI without their permission. This includes giving creators control over the price and value of their work.

In my view, this is a revolutionary step in right direction and could represent one of the strongest commitments to copyright protection in the AI era made by any government so far. Protecting creative work is essential to preserving innovation, encouraging artistic expression, and ensuring that creators are fairly recognised and compensated for their contributions. This will encourage greater public confidence in AI, which is essential for its long-term growth and continued innovation.

It is reassuring to see these concerns acknowledged at the highest level. I hope these commitments are translated into effective legislation in Australia and, ultimately, inspire similar protections in other countries as AI continues to evolve."

“It’s gratifying to see a level of responsiveness to the advances and the risks of AI coming from the federal government and to see a move to binding protections and guardrails.

Beyond the form and ownership of this oversight body, we need to hear much more about the content and how they will prioritise and address specific ethical issues.

Beyond those mentioned by the PM today such as data centres and copyright protections, and beyond general talk of ‘Australian values’, we should be prioritising the safety and protection Australians online, especially young Australians, as well as threats to our democracy from misinformation and bad actors.

All of this should be bolstered by meaningful platforms and opportunities for well-informed public discussion, so the public can have the opportunity to understand these opportunities and risks and have their voices heard as part of well-informed and robust public dialogue on these rapidly emerging issues. Our public universities also need to play a strong role here.”

"Australia has taken an ad hoc, case-by-case approach to AI, particularly when it comes to data and data centres. As a result, significant amounts of Australians' personal data are being sent overseas, where they may be processed, stored, or both, often with limited transparency.

Australians deserve greater oversight of how their data is handled. We need clear safeguards around what data is being collected, why it is being collected, who is collecting it, and where it is ultimately stored. These are fundamental questions that go to the heart of privacy, security and personal identity.

I am happy to see the government recognising these challenges and taking steps to introduce regulation in this sector."

"Artificial intelligence is a general term that covers multiple technologies that allow for automation, generation or rapid calculations. It spans using generative AI to create a funny image to using medical technologies to review scans with potential for life saving outcomes. Understanding the different forms of AI is crucial to discerning the appropriate applications of these technologies to the needs and values of Australians. Clearly explaining this to the Australian public will be crucial so that they can determine how these various AI technologies will impact everyday life for better or worse.

Because AI isn’t one singular technology, this means that different sectors will need to respond differently. While the establishment of the Office of AI is beneficial by coordinating these conversations in one space, it is still difficult to see how competing needs will be dealt with.  For instance, by committing to placing data centres outside of urban areas, there emerges conversations of what land is ecologically or culturally appropriate for data centre development.

Senior level support and a central location are key indicators of success for a new government office. The Office of AI has these two core ingredients by being placed in the Department of Prime Minister and Cabinet."

"I am encouraged by the whole-of-nation response to the fast-changing developments in the AI space. This is the most effective way to increase the accountability of AI technology companies as they interact with the data and lives of all Australians. The establishment of the Office of AI also offers unique opportunities for Australia to think about how our nation can update our regulations to pace with AI, and invest in research towards key challenges such as provenance of AI, data centre energy efficiency and many other challenges such as the increased usage of AI for cyber attacks against Australian enterprises."

"Today's announcement was really timely, but also partially overdue because we did miss the opportunity to be at the forefront of innovation with AI, and only now we're just exploring data centres.

There have been three years of Ed Husic's AI regulation inquiry, which I also contributed to as part of my previous research centre, and that didn't result in any actionable policy or regulation outcome - which resulted in businesses being reluctant to take up AI. For example, a recent survey by KPMG shows that Australia ranked sixth from the bottom out of 46 countries in trust in AI, and that's led to the need for that social licence that Albanese has been talking about, because we have been focusing on the threats of AI and withholding the regulations, leading to business uncertainty and slowing adoption.

Now what is concerning is that the data centres that prompted another look at AI pose a threat of potentially running out of electricity and energy, and that's not something that we can provide because we're trying to transition to green energy, so it's our bottleneck. Data centres already consume around 3–5 per cent of Australia’s electricity, and AEMO expects this to rise to between 6% and 12%.

On the other hand, Albanese also mentioned that the problems with AI are regulatory, social, and economic, and while there's been a lot of talk about the threat of AI to jobs, that's actually not something that businesses see as a problem. Australian data shows that 81% of business executives have seen no impact on employment over the past three years, and 67% of them don't foresee any negative impact in the future. So we are technically more likely to have our jobs lost to overseas workers than to AI.

To sum up, the announcement is somewhat overdue, but it's still needed and needed urgently to provide the clear regulatory pathway - more clear than the existing AI framework - and more clear than what has been provided by the Productivity Commission in one of the five pillars last year from the roundtable that they conducted."

"Prime Minister Albanese has outlined an AI vision that touches upon regulation, ethics, and a possible mandatory national AI framework. The latter represents a significant shift from the Federal government’s earlier position where there would be no separate standalone AI Act.

Aside from some very short-term opportunities for private sector investment in data centres, the PM’s announcement did not focus on how the government will capitalise long-term on the AI opportunity for significant investment from both the private and public sectors. This is contrary to the approach of both the UK and Canadian governments, whereby their AI strategies have seen concrete financial imperatives (in partnership) to leverage the AI opportunity for the benefit of their respective countries."

"Clearly, the government wants to unify its response and the laws surrounding AI. It is encouraging to hear that the focus is not on stopping or pausing AI, but on working out how it can best serve Australians.

The strongest aspect for me was the statement that Australian creatives and journalists should not have their work used without permission as training data, and that they should retain direct control over how their work is used. This is in stark contrast to the United States and appears to lean towards an even stronger position than that of the European Union.

Australia is well placed to move forward, with many positive attributes, from being one of the sunniest nations on Earth to having a stable democracy. However, what was not addressed was increased funding for R&D, basic research, or universities.

The intent and messaging suggested that Australia should not simply be a user of AI or a landlord for overseas data centres, but an active participant in its development. For this to happen, we need greater investment in science and for industry to rebalance its funding towards long-term innovation.

For context, here is one of the major AI projects in the Motus Lab at the University of Sydney."

Australia must invest in AI skills and training to strengthen cyber security, improve productivity and prepare businesses for emerging risks.

Small and medium-sized businesses are often more exposed to cyber attacks but may lack the resources to respond. Larger organisations, particularly those with experience managing cyber incidents, should share expertise, training and practical support.

AI can improve decision-making by analysing large amounts of data, identifying patterns and detecting risks earlier. However, it should support rather than replace human judgment. Poor-quality data, biased systems and over-reliance on automated outputs can lead to inaccurate decisions, unfair outcomes and new privacy or security risks.

Employees at every level need clear guidance on using AI safely and responsibly. Business leaders must understand its benefits and limitations so they can question AI-generated advice and remain accountable for final decisions.

Australia must also build public trust through AI education in schools, vocational training and universities.

Further research is needed to assess the reliability of grid connections for data centres, and their effectiveness should be tested before they are introduced at full scale.

"The establishment of the Office of AI is an important step. It shows that the government no longer sees AI as just a technology or industry issue. Instead, it recognises AI as a whole-of-government challenge that cuts across productivity, national security, jobs, energy, local IP, and social stability, and one that requires leadership from the centre of government.

At the same time, the announcement highlights a key challenge. While the government recognises that AI will have far-reaching impacts, it has not yet developed a complete national strategy that explains who should own and control key AI capabilities, how the benefits of AI should be shared across society, and what kinds of overseas dependence are acceptable.

The most important question is therefore not simply whether an Office of AI should exist, but whose AI strategy it will serve. Will it mainly make it easier for global technology companies to invest in Australia? Will it focus on using AI to make the public service more efficient? Or will it develop a longer-term strategy for building Australia’s own capabilities and protecting the broader public interest?"

"The establishment of an Office of AI and a national framework is a significant and necessary step. It is important to be clear about what AI is and what it is not. AI is not magic, it is mathematics, built on statistical models, data, and computational scale. That matters because its limitations are measurable, its risks are quantifiable, and its performance depends on design, data quality, and oversight.

The Prime Minister is right to frame this as a generational opportunity. AI can be a powerful tool to enhance productivity, support decision-making, and strengthen national capability. However, that opportunity will only be realised if Australia invests not just in infrastructure and adoption, but in foundational capability. World-class AI depends on deep expertise in mathematics, statistics, and decision science, and these skills must be actively developed and supported.

AI should be used to amplify human expertise, not replace it. Building and maintaining that expertise is essential if we are to design, evaluate, and challenge these systems effectively, rather than becoming dependent on them.

The emphasis on resilience is therefore critical. Sovereign capability in AI is not only about infrastructure or regulation, but about ensuring we have the knowledge and capability to understand and improve the systems we rely on."

"The creation of an Office of AI is a welcome and overdue step. AI is no longer a narrow technology issue. It affects the economy, government services, national security, education, jobs and the creative sector. Placing the Office within the Department of the Prime Minister and Cabinet should give it the reach needed to coordinate policy across government.

Of course, creating an office is only the beginning. Its value will depend on whether it has genuine technical expertise, clear authority and the ability to hold agencies and companies accountable. Australia needs rules that provide certainty and protection without becoming outdated as the technology develops.

The Prime Minister is also right to reject the idea that innovation requires weakening copyright. Australian writers, artists, journalists and other creators should know when and how their work is being used. AI companies should pursue fair licensing arrangements rather than assume that everything available online is free to exploit.

We should take the same measured approach to data centres. They can bring investment and strengthen Australia’s digital capability, but they also consume significant electricity and water. Faster approvals must not come at the expense of proper environmental and community scrutiny.

Ultimately, Australia should aim to build its own AI capability, not simply become a customer for technologies developed elsewhere."

"The establishment of the Office of AI is an important step. It shows that the government no longer sees AI as just a technology or industry issue. Instead, it recognises AI as a whole-of-government challenge that cuts across productivity, national security, jobs, energy, local IP, and social stability, and one that requires leadership from the centre of government.

At the same time, the announcement highlights a key challenge. While the government recognises that AI will have far-reaching impacts, it has not yet developed a complete national strategy that explains who should own and control key AI capabilities, how the benefits of AI should be shared across society, and what kinds of overseas dependence are acceptable.

The most important question is therefore not simply whether an Office of AI should exist, but whose AI strategy it will serve. Will it mainly make it easier for global technology companies to invest in Australia? Will it focus on using AI to make the public service more efficient? Or will it develop a longer-term strategy for building Australia’s own capabilities and protecting the broader public interest?"

"The Australian government’s announcement of a new, central Office of AI, to sit in the Department of Prime Minister and Cabinet, is a welcome move that will provide clarity and guidance for businesses and communities. The Office will coordinate work across government departments and create new Australian standards for AI adoption.

As technology companies increasingly look to build data centres in Australia, communities have expressed significant concerns about energy and water use, and placement near residential neighbourhoods. Australians are also worried about disruption from AI technologies, raising concerns about potential job losses, copyright infringements of material used to train AI models, and data privacy.

Addressing such varied challenges, particularly when AI technologies are evolving at such a fast pace, warrants the coordinated approach this office will take. Given the known risks posed by generative AI technologies, for example, from deepfake images that mislead customers, to hallucinated content that misinforms chatbot users, Australians need government intervention to protect them from harm and ensure AI use is in their best interests.

The creation of this office marks a significant shift in the government’s overall approach to governing AI, towards being more hands-on and proactive. In 2024, the government planned to regulate high-risk AI implementation but later abandoned that plan. The government also proposed - and then paused - digital duty of care legislation, to hold technology companies to account for harmful content and tool designs. By creating this office (and by saying digital duty of care legislation will be reintroduced this year) the government has recognised the need for coordination and intervention to address the varied risks posed by AI, while also supporting its potential benefits.”

“Artificial Intelligence is going to change the way we work and the underlying drivers of the economy. It’s critical that Australia be at the forefront of the AI revolution and that our workforce is trained and prepared for this new wave of technology.

The critical balance with this new AI framework is ensuring the safe and responsible use of the technology whilst not stifling innovation. It is essential that Australia has a strong sovereign capability in AI without relying on overseas partners for a capability that will become as important as access to power and telecommunications.”

"The Prime Minister’s instincts on AI are on point. Critically, he understands that “we should not treat AI as a threat to good jobs, we must use it as an instrument to help create them”.

Australia can’t hold back the tide, but we can surf it, and we can even redirect it.

Doing it the Australian way, true to our values and standards is the right approach. His focus on the energy and water needs for data centres is essential. And most importantly, he’s now fired the starting gun on a process to consult across government.

For universities and scientists, that will particularly mean with the education minister and the science, technology, and digital innovation deputy minister.

AI is revolutionalising research and teaching, not always in ways that are well understood or welcome. So, it’s very good that the whole of government is now alert to this change and their role in ensuring AI works for Australians."

"The establishment of an Australian Office of AI and a national AI framework represent a significant step towards Australia's approach to AI future.

Two things would be critical:

1. It will be important to adopt a balanced view that recognises both the opportunities and the potential risks/cautions associated with AI. Its success should not be assessed through economic outcomes alone. Equal attention must be given to how AI affects people’s experience of work, communities, inclusion, and wellbeing.

2. A collaborative and transparent approach will be critical. The national framework should be developed through meaningful engagement with industry, unions, community representatives, academics, and regional stakeholders. This will help ensure that Australia’s approach to AI is practical, inclusive, and equitable, to the ways in which AI may affect people, organisations, and communities."

"Establishing an Office of AI within Prime Minister and Cabinet is the right thing to do, but making it work will be easier said than done. Until now we have approached AI governance in a piecemeal fashion with departmental initiatives and voluntary guidelines. Having a unified national framework would give business and researchers the regulatory basis that they can proceed with. This would allow Australia to proactively shape the technology rather than simply using it as supplied.

Declining to give AI firms a blanket waiver to use Australian data is a good move that will protect creators of digital assets. But the government will need to define what lawful access looks like and that's not so easy. If our licensing pathways are difficult to use because they are unclear or convoluted then the developers will go offshore and Australia will be left to import the finished product but without any of the economic benefit.

The test for the new Office of AI will be whether it can find the 'sweet spot' between the risks and the real cost of moving too slowly. Regulation that arrives late or lands heavily will not protect Australians. All it will do is make sure the technology that shapes our lives will be built elsewhere, under someone else's rules."

“This announcement is too little, too late. The  Australian government is finally waking up to the AI reality, but it continues to approach the challenge primarily as one of attracting investment with minimal risk mitigation.

We have seen this model before in the gas industry. Foreign corporations use Australian resources, infrastructure, and labour, export much of the value, and leave Australians carrying substantial environmental and social costs. AI risks repeating and amplifying that mistake. Its data centres consume Australian land, water, and electricity, while its models draw on our research, creative work, data, and intellectual labour. Yet ownership, control, and profits remain concentrated in a handful of foreign corporations and their executives.

The gas analogy, however, only goes so far. AI is not merely another extractive industry. It increasingly shapes how Australians work, learn, communicate, relate, and access or produce vital information. Dependence on privately controlled AI infrastructure is therefore a question of democratic and national sovereignty, not only taxation or regulation.

Rejecting a copyright free-for-all is necessary, but far from sufficient. Australia needs collective sovereignty over AI infrastructure through meaningful public and community ownership, genuine citizen control, and a fair distribution of benefits according to social need. Otherwise, we will once again socialise the costs while privatising the gains.”

"The establishment of the Office of AI is a welcome step that elevates AI to a whole-of-government strategic priority and should improve policy coordination across Australia.

National AI standards are also positive, noting that they need to create guardrails that foster trust without restricting or hurting innovation or investment.

The announcement seems to strike a reasonable balance between promoting AI innovation and managing its risks, avoiding an overly risk-averse approach that could leave Australia behind.

One notable gap in the announcement is AI literacy. Australia needs a national capability-building strategy, educating everyone from school students to organisational leaders and policymakers to thrive in an AI-enabled future while being mindful of risks."

"The creation of this office is timely, considering how quickly AI technology is advancing. It should help address governance challenges and work alongside the government's cybersecurity efforts."

"The Office of AI is the right instrument. Putting it in Prime Minister and Cabinet is the right call, because AI is not an industry portfolio, it is a whole-of-society one. The question now is what sits inside it. Standards design covers the technology. Nothing yet covers the people. An Office of AI without a human transition function is an office for managing machines in a country made of humans. Australia now has a body responsible for coordinating how AI is built here. We still don't have one responsible for how Australians get through it.

The Prime Minister is right that the world should have acted on social media a decade earlier. But look closely at what we actually got wrong. We didn't get the infrastructure wrong. We got the psychology wrong. The harms were developmental, cognitive and social, and they arrived years before we had the evidence or the institutions to answer them. If we take that lesson seriously, we don't just build better rules for data centres. We build the capability to see human harm coming before it shows up in a statistic. The social media reckoning was not a failure of engineering standards. It was a failure to understand minds. We are about to make the same category error at a much larger scale.

Today we learned where data centres can be built, what power they must generate, how much water they may use, and who owns a song. Every one of those is a rule about the technology. Not one is a plan for the people the technology is about to move through. We are regulating the footprint and ignoring the population standing in it.

The Prime Minister noted that extremists and state actors already use AI to target young people with propaganda. Every dollar of that threat lands in a human mind. You cannot firewall a population. The defence against synthetic persuasion is a citizenry equipped to recognise it, and that is a psychological capability, not a technical one. It is currently nobody's job to build it.

We want AI to support and create good jobs, not replace them.' Everyone wants that. Wanting is not a mechanism. Data centres now have a legal obligation to put more power into the grid than they take out. That's a mechanism. There is no equivalent obligation, anywhere in this framework, to give a single Australian worker warning, retraining, or support before their role is automated."

"AI sovereignty will be critical to ensuring AI in healthcare develops on Australia’s terms, not those of overseas technology providers.

The new national policy roadmap for AI in healthcare (to be released 22 July by the Australian Alliance for AI in Healthcare) - developed through extensive consultation with more than 96 organisations across government, industry, healthcare, consumer groups and peak bodies - argues that healthcare requires special attention because of its complexity, sensitivity and reliance on trusted patient data.

Without sovereign AI capabilities, Australia risks exporting valuable health data and importing AI systems that may not reflect local populations, services or needs.

The roadmap calls for local governance, infrastructure and expertise to reduce dependence on international providers and protect Australians from emerging risks. Its message is clear: the pace of AI change is accelerating - this is not the time to walk, but to run.”

“I welcome this important initiative. Developing sovereign AI capabilities is critical to Australia’s future, and the establishment of the Office of AI represents an important first step.

While many details are still to emerge, it will be vital to bring together Australia’s leading research capabilities and industry expertise to help shape the Office’s direction.

A coordinated and collaborative approach will be essential to positioning Australia at the forefront of AI-driven innovation.”

“This is an important step. AI technology will be a foundation for Australia’s future and this initiative acknowledges the key role it will play.

I see the area of AI safety as being primary in future discussions - having the capacity to assess the quality, safety and reliability of AI is critical for future adoption.”

"Australia's decision to refuse a copyright waiver for AI firms is the most significant part of today's announcement, and it deserves more attention than it is getting.

For years, writers, journalists, musicians, and academics have watched their work absorbed into AI training datasets without permission, payment, or even acknowledgement. The PM's statement that "anything less is theft" is a remarkable thing to hear from a sitting prime minister, and it reflects a growing recognition that the creative economy cannot be the unpaid foundation of a commercial AI industry.

From my perspective working with students in media and communications, this matters enormously. We teach students to value their creative output, to understand their intellectual property, and to build sustainable careers from their skills. A national framework that allows AI firms to mine Australian content without consequence sends entirely the wrong message to that next generation of creators.

The copyright commitment also signals something about the kind of AI future Australia wants. Rather than racing to attract investment at any cost, this government appears willing to draw a line around what belongs to Australians. Whether that commitment survives the lobbying pressure that will inevitably follow is the question worth watching."

"I’ve been saying we need government wide handling of AI since ACOLA report in 2019. It’s an economy wide change that needs a whole of government response.  So I welcome the PM’s attention, the establishment of the Office of AI, especially that it will sit within PM&C, the centre of government.

When the UK announced its Office of AI in 2019, they also announced GBP 1 billion of additional funding, the so called 'AI sector deal' to get some of the upside as well as avoid downside. I would love to see Australia look to seize some of the upside not just avoid downside. Are we going to see the government invest in our AI research base?

But broadly, the PM touched on many of the issues justifiably troubling the public such as data centres in their backyard, AI taking jobs, copyright and creative industries, the impact of AI on our kids.

The PM understands the issues. The question now is whether this new Office will address them adequately."