AI is creating a new strategic infrastructure race
The exponential growth of AI, and the infrastructure needed to support it, are reshaping the world.
It’s hard to think of an area of life in which AI technology is not playing a role: work, study, business, investment and government, through to recreation and daily life.
“To ask ‘whether AI is a bubble’ is a foolish question. AI will transform our lives completely, and do so in a way that generates profits. Those who refuse to evolve are closing down their world. Those who condemn AI are themselves spitting upward.”
Value in Motion
How to think like an AI revolutionary and unlock value in the decade ahead.
For more
The rise of AI is already having major impacts on capital markets, energy and resources needs and industrial, social and governmental systems across the world.
In this report we explore ways in which Australia and Japan can work together to leverage combined strengths and capitalise on the AI megatrend.
While the US and China are leading the way in the AI race, Japan is a vital part of the AI value chain through its technological and manufacturing strength in the world of chips and semi-conductors.
Japan, though, faces constraints in domestic land, energy security and sovereign compute capacity. In this respect, it has a natural partner in Australia.
Australia has many of the inputs needed to host key components of AI infrastructure: land, energy resources, critical minerals, political stability and trusted governance.
The technology sector is now the second largest contributor to Australian GDP.2 Australians are early adopters of all forms of new technology and can be expected to embrace the productivity-enhancing advantages AI offers. The challenge for Australia will be to deepen its involvement across the AI ecosystem to capture the gains that follow.
“The question is not whether AI lifts productivity. It will. The question is who captures the gains. If Australians own the tools, we share in the upside, higher incomes, stronger national wealth, broader opportunity. If the tools are owned elsewhere, and we simply rent them, the dividends flow offshore. Productivity may rise, but national income does not necessarily follow. We will gain some benefits from adoption, but relative to other countries we may fall behind.”
“Aligning government policy, investor confidence, energy strategy, planning systems and industry capability is what will ultimately determine whether Australia captures this moment and maximises the benefit,” PwC’s Global Infrastructure Leader Clara Cutajar says, in response to the Australian government’s guidance on its expectations for data centres and AI infrastructure developers.
Every opportunity comes with constraints, and AI infrastructure is no exception, with legitimate concerns ranging from energy demand and water use to community acceptance of data centres. Recognising these considerations, this report focuses on how AI and its supporting infrastructure could nevertheless open a significant new area of collaboration between Australia and Japan.
Australia may not be the lowest cost market, but it can offer compelling risk-adjusted returns and a high level of strategic and geopolitical alignment for Japanese companies, many of which have a long and successful history of operating in Australia.
Australia and Japan have the opportunity to collaborate on AI infrastructure and create the building blocks to export compute capacity on demand to Asia.
The opportunity is to extend the historic Australia-Japan partnership into a new “energy-minerals-compute” partnership for the AI age.
As defence partners and allies of the US, the two nations also have the opportunity to collaborate on AI systems for military and intelligence uses.
Sovereign AI needs and the closeness of the Australia-Japan partnership offer a way forward with Japan well placed to supply Australia with technology to drive its AI industry and receive in return extra secure compute capacity from a trusted partner.
The AI infrastructure revolution
Al investment is occuring at an unprecedented scale globally, with the world’s five largest technology companies expected to commit more than US$1 trillion in AI-related capital expenditure across 2025 and 2026, helping to create a major new infrastructure asset class.4
While public attention focuses on AI models such as ChatGPT and Claude, the broader AI value chain extends from energy and critical minerals through semi-conductor manufacturing, servers, data centres and associated power and cooling infrastructure to the applications used by businesses and consumers.
“As AI places new demands on our infrastructure systems, the focus must be on co-ordinated investment that unlocks innovation and growth at scale. If we get that right, infrastructure can become a powerful driver of long-term national prosperity."
AI is effectively an interconnected technology-energy-capital system: data-centre demand is translated into usable compute through semi-conductors, servers, networking, power generation, grids, cooling, water, fibre, financing, permissions and operating capability.
Reflecting the scale of the associated investment, the latest PwC Global Data Centre Outlook 2025-50 estimates that global data-centre capital expenditure could reach US$31.6 trillion through 2050, rising towards US$50 trillion if AI adoption accelerates.
“The data centre is essentially a chip-replacement subscription with a building wrapped around it. Build the box once, refill it every four to six years. That pattern has implications not just for how much capital is needed, but for who is servicing the demand it meets”
Crucially, this is not a conventional infrastructure cycle in which spending peaks once construction is complete. Much of the capital will be directed towards servers, GPUs and other equipment that must be replaced repeatedly as technology advances.
PwC’s Data centres at the crossroads of technology and resilience report, released last year notes that AI’s success depends on access to power, with the International Energy Agency expecting data centre electricity demand to double by 2030.
Increasingly, major tech firms are signing renewable power agreements and building their own generation capacity, making them major investors in wind, solar and nuclear. Amazon has invested over A$2.8 billion in 20 renewable energy projects in Australia since 2020, illustrating the scale of its commitment to clean energy.5
Critical minerals - lithium, cobalt, nickel, rare earths, copper, silver, aluminium, silicon and gallium - are just as vital to AI's expansion as energy and nations will pay increasing attention to securing supply.
Silver and copper already face projected supply deficits in an era where bringing new mines online can take more than 20 years.
Many of these minerals remain highly concentrated geographically, creating a strategic incentive for trusted supply-chain partnerships.
Global Data Centre Outlook 2026-50
Where $31.6 trillion of capex flows in the era-defining AI build-out
Global data centre capital expenditure could reach US$31.6 trillion through 2050, with a plausible upside of nearly $50 trillion if AI adoption accelerates.
Australia and Japan in the global AI age
Source: Cabinet of Japan and Nikkei Asia6
Japan has rapidly increased its investment in AI and data centres at a corporate and government level in recent years, underscoring its ambition to become a significant player.
The Takaichi government's Honebuto no Hoshin ("big-boned policy") aims to plough 370 trillion yen of public and private funds (US$2.4tn) into priority sectors including AI, robotics, semi-conductors and energy by 2040. The policy is driven partly by concerns over inadequate domestic AI computing capacity and heavy reliance on American cloud providers (AWS, Microsoft, Google, Oracle, IBM).
Most of Japan's AI infrastructure is clustered around greater Tokyo and Osaka where Australian companies, including Goodman Group, NEXTDC and AirTrunk have made recent investments. However, as land and energy constraints bite, capital and policy are pushing data centre development towards secondary hubs such as Hokkaido, Kyushu, Chiba, Saitama, Nagoya, Yokohama and Fukuoka.
“In the Asia-Pacific region, we see four areas emerging as the principal cloud and AI hubs: Japan, Australia, Singapore–Malaysia and India. Japan is a strategic market that will support not only domestic demand but North Asia as a whole.”
Australia has attracted foreign AI capital largely from North America, with Sydney and Melbourne the major geographic focus — though projects in Queensland (including the recently announced $30 billion Zerra DC/Macquarie Capital data centre northwest of Brisbane to service Claude developer Anthropic), Tasmania, South Australia, Western Australia and the ACT are expected to come on stream.
Elevated construction costs, changing regulatory procedures, energy policy uncertainty and community opposition are among the current challenges facing data centre developments in Australia, pressures that are likely to intensify as the sector’s electricity use increases. For instance, data centres consumed 2% of electricity in the National Electricity Market (NEM) in 2024–25 and the Australian Energy Market Operator (AEMO) expects this demand to almost triple to 6% of NEM electricity consumption by 2030, equivalent to the power used by all Victorian homes.8
The Australian government has proposed new standards mandating that new data centres do not add to overall power demand or drain water resources. But the flipside of tighter regulation is a clear, consistent national stance that can underpin investor confidence and build community acceptance over time.
The challenge for Australia and Japan will be to compete with the wider world for AI capital in a way that plays to their complementary strengths and creates a win-win outcome.
Strategic areas for Australia-Japan collaboration
The AI value chain is about much more than models and software.
It starts with the physical foundations, namely energy, water, critical minerals and land that make it possible to manufacture semiconductors, build and operate data centres, and provide the computing power on which AI depends. Those foundations ultimately support the AI tools and applications used across business, government and society.
Australia and Japan bring different but highly complementary strengths to this value chain.
Australia has abundant land, energy and critical minerals, while Japan brings capital, advanced semiconductor and manufacturing capabilities, and deep industrial expertise.
By combining these strengths, the two countries could build a shared AI infrastructure and compute platform that serves their own needs and supports wider demand across the region.
The opportunity builds step by step: resources support hardware; hardware provides compute; compute powers AI models; and those models enable applications that can lift productivity, innovation and strategic capability.
Against this backdrop, the four connected areas where Australia-Japan collaboration could have the greatest impact are: land and energy; critical minerals; AI infrastructure and data centres; and sovereign compute and AI applications.
Taken together, they could open a new chapter in the long-standing relationship: an energy-minerals-compute partnership for the AI age.
Land and energy
Data centres require significant space in areas where community support and energy supply can be secured. Larger campuses have a footprint of more than 500 hectares and having additional space provides more options for cooling and power.
Outside Hokkaido, with its wide-open spaces and abundant water and energy, this combination is hard to find in Japan.
Australia, by contrast, has abundant developable land for large-scale campuses with room for long-term expansion and integrated energy infrastructure. As land constraints intensify in established global hubs, this gives Australia a structural advantage in attracting next-generation AI investment.
On energy, Australia holds significant reserves, both renewable and hydrocarbon based, though energy market and policy settings remain a significant consideration and the current supply/demand balance is pushing prices higher. Coal plants are gradually retiring and new supply is mostly in the form of intermittent renewables.
With policy moving to ensure that data centre developments do not add to aggregate demand, there will be a push to unlock new supply. The same process is playing out across the world and leading to a revival of LNG generation, along with increased nuclear generation in countries that allow it.
Expanding LNG power generation to power the AI infrastructure boom is a genuinely fertile area for collaboration as Australia has ample reserves while Japan has long provided the capital and offtake agreements needed to develop major projects. Sustaining this collaboration, however, will require ongoing dialogue to align Australia’s domestic energy security priorities with the investment certainty Japanese partners need to commit at scale.
Japanese utilities and trading houses are seasoned investors in Australia’s resources sector, having helped build the nation's iron ore, coal and gas industries, and can be relied on to fund AI-driven expansion, aided by deep familiarity with Australian law and business conditions.
For instance, Mitsui, already active in Australia’s energy sector, has unveiled a global strategy to expand LNG investment across Australia, the Middle East and the US to capitalise on rising data-centre power demand. “Without securing energy, it is impossible to implement solutions,” Mitsui & Co CEO Kenichi Hori told Bloomberg when outlining the company’s positioning to meet this growing global need.9
Critical minerals
Critical minerals offer another clear opportunity for Australia-Japan collaboration. Australia has significant reserves of many of the minerals that will underpin the global AI infrastructure build-out, though their exploitation has been held back by a lack of capital, and in some cases, Chinese influence on supply and pricing.
“It’s no longer enough to say Australia has a first-class natural endowment of critical minerals. Our relative abundance is proving challenging to convert. And projects are not progressing through the pipeline quickly enough. That is fertile ground for bilateral collaboration with Japan, ensuring a timely and secure supply of the critical inputs for the next generation of AI infrastructure."
Lynas, with its Mt Weld mine and Western Australia concentration plant, plus a Malaysian processing plant, is one of Australia's leading rare earths producers, now valued at $A16 billion. It's a strong example of Australian-Japanese collaboration: Sojitz provided long-term finance and technical support throughout the project's turbulent history, including when Chinese supply surges drove prices lower.
Australia is actively working with Japan, South Korea, the US and others to build an ex-China supply chain for critical metals.
Australia holds significant reserves of a range of technology metals including:
- Rare earths: significant light and heavy rare earth deposits, with several players looking to follow Lynas' lead
- Copper: world's second-largest reserves (behind Chile); South Australia's Olympic Dam operation produces 300,000 tonnes annually
- Gallium: recoverable as a bauxite/alumina by-product from the world's second-largest bauxite reserves; the Alcoa Gallium Recovery Project (backed by Sojitz and Japan Organization for Metals and Energy Security) is on track to supply up to 10% of global production
- Lithium: a top global producer, regularly contributing over 40% of world output
- Nickel/Cobalt: second-largest global reserves of both (behind Indonesia and the Democratic Republic of Congo respectively); the Ardea Resources Kalgoorlie project is a joint venture with Sumitomo Metal Mining and Mitsubishi Corporation.
Aussie Mine Series 2026
Can we get serious now?
Securing Australia’s critical minerals future
Australia has the preconditions for critical minerals success. 900+ critical minerals projects in the national pipeline speaks to the unprecedented opportunity. Yet only 13 of those projects are within striking distance of a Final Investment Decision (FID) in the next 2–4 years.
Source: US Geological Survey, January 202610
AI itself can accelerate this advantage, discovering and mapping reserves more efficiently, and reinforcing Australia's position as a minerals and energy supplier for the AI age.
AI infrastructure and data centres
While Japan has diversified its foreign investment into Australia across multiple industry sectors, including in AI-adjacent technology, it has yet to move at comparable scale into the physical data centre and compute space. That gap represents one of the clearest opportunities for the next phase of Australia-Japan economic collaboration.
Recent transactions demonstrate that Japanese investors understand the Australian technology market and have an appetite to do more. Mitsubishi Electric's 11.5% stake (A$52.5m) in Australian driver-monitoring AI company Seeing Machines, gave Japan exposure to Australian AI technology. Meanwhile, Renesas Electronics' A$9.1 billion acquisition of Altium, one of the largest Japanese acquisitions of an Australian company in any sector, brought together Japanese semi-conductor capability and Australian electronics-design software.
“The next five to ten years will determine Australia’s infrastructure trajectory for the next quarter-century … Capital alone won’t set us apart. Alignment will, across policy, planning, funding, delivery and community outcomes. If Australia can build this alignment now, infrastructure can do more than just keep pace with growth. It can strengthen productivity, accelerate decarbonisation, improve resilience and help shape the next era of national prosperity.”
Japan has yet to bring that appetite to the physical data centre and compute layer itself, where the capital being deployed dwarfs even the Altium deal (AirTrunk alone was valued at A$24 billion when Blackstone and CPP Investments acquired it in September 2024).
Japanese AI-related capital has so far gone overwhelmingly to the US, India and Southeast Asia instead with the US offering an established ecosystem Australia can't yet match, and India and Southeast Asia offering lower costs and faster approvals. But Australia can compete through renewable energy potential, available land and geostrategic alignment with Japan.
Australia is forecast to invest around A$6.2 trillion in infrastructure by 2050, including approximately A$449 billion in digital infrastructure, according to PwC Australia’s Infrastructure to 2050 report. But scale alone will not make Australia a regional AI hub for Japan. Its competitive advantage will depend on co-ordinating power, grid connections, land, planning, water and digital networks, and providing investors with the speed and certainty required to commit capital.
One recent validation of Australia’s appeal was Tokyo-based Datasection deploying a large-scale GPU cluster in DigiCo's SYD1 centre in Sydney to serve Asia Pacific demand thought to be linked to Tencent, starting with 20 megawatts, double the capacity planned for its Tokyo and Bangkok projects, and scaling to 60 megawatts within two years.
Chinese giant Tencent is a huge consumer of compute capacity through offering streaming, gaming and social messaging services to more than a billion users throughout Asia and beyond.
This investment suggests Australia can serve regional customers better from a latency advantage compared to facilities based in the US.
“For emerging AI and cloud operations, Australia offers several advantages: comparatively abundant land relative to dense Asian cities, a stable policy environment, strong renewable energy generation and the potential for competitive energy costs as the grid transitions. These factors support the case for expansion, though the country faces competition to become an Asia‑Pacific hub”
If Australia can convert this early interest into a repeatable investment model, it could become a trusted regional platform through which Japanese capital and technology serve Asia-Pacific demand. The prize is not simply more data centres, but a deeper bilateral partnership spanning energy, digital infrastructure and compute.
For Japanese investors, the opportunity is therefore substantial, but its attractiveness will depend on whether Australia can turn infrastructure scale into an integrated, investable proposition.
Sovereign compute and AI applications
Beyond commercial compute, there's potential for collaboration on "sovereign" compute capacity leveraging Australia and Japan's tight security relationship.
Both countries are close partners and allies of the US and collaborate on defence and intelligence sharing and have recently struck economic and cyber security agreements that extend to partnering on AI.
Combining on AI systems in defence and national security that are reliant on trust and shared values is a promising proposition for the two nations.
As the Australian Strategic Policy Institute's David Wroe put it, "the government should lean into this opportunity as far as Japan is willing to go", Japan being "a regional neighbour whose strategic outlook is the closest to Australia's of any Asian nation and that has proven... it can move technological and industrial mountains.11"
The growing use of AI in defence and security creates scope for Australia and Japan to develop a trilateral compute partnership with the US. Within this arrangement, Australia could provide secure AI infrastructure for less latency-sensitive workloads, including AI model training, supplementing Japan’s domestic capacity.
Conclusion
AI is creating a new layer of strategic infrastructure in the global economy, one that will require an estimated US$31.6 trillion of data-centre capital expenditure by 2050 and reshape how countries compete for energy, critical minerals and computing capacity.
Across the four areas considered in this report: land and energy; critical minerals; AI infrastructure and data centres and sovereign compute, Australia and Japan bring together a combination of strengths that few other partnerships can match.
This kind of complementarity is not new.
"The next decade of AI infrastructure will be defined by the partnerships nations choose to build now. For Japan, deepening our engagement with Australia across energy, critical minerals and compute is not simply a commercial opportunity. It is a foundation for the technological sovereignty we will need in the AI age."
For nearly 70 years, the Australia-Japan relationship has grown by matching the strengths of one country with the needs of the other. Australian wool supported Japan’s textile industry. Iron ore and coking coal helped produce Japanese steel. And LNG has helped power Japanese homes and industry.
AI could be the next chapter in that story, taking the same partnership model further up the value chain and into the compute economy. And the benefits could flow both ways.
AI infrastructure powered by Australian energy and critical minerals, could support systems that help discover new deposits, run processing plants more efficiently and modernise the electricity grid, strengthening the industries that made the infrastructure possible in the first place.
The geopolitical environment adds to the opportunity.
As the world becomes more multipolar, with China playing a larger role, the US arguably stepping back from its traditional leadership position in the Asia-Pacific, and demand growing for trusted supply chains, the Australia-Japan partnership is becoming more strategically important.
The 2026 Joint Statement on Elevated Critical Minerals Cooperation, the Australia-Japan Joint Declaration on Economic Security Cooperation, and the Australia-Japan Strategic Cyber Partnership signal that the two governments are already moving in the same direction. Capital and industry now need to turn that alignment into co-ordinated execution.
The next five to 10 years will shape the trajectory of AI infrastructure investment across the Asia-Pacific for a generation.
The foundations are already in place for Australia and Japan to expand their investment and collaboration in this vital and rapidly growing sector.
If Australia and Japan can convert today's alignment into a repeatable, investable partnership across energy, critical minerals, digital infrastructure and sovereign compute, both countries stand to gain, not just in economic terms, but in the security, resilience and technological sovereignty that will define the AI age.
Bridging Continents
Inbound investment from Asia to Australia
For more
Contact us
