The Federal Government deserves credit for recognising that artificial intelligence has become a matter of national policy. Its new AI strategy addresses important issues such as governance, energy consumption, water use and copyright protection.
These are sensible initiatives.
But these are not the issues that will determine Australia’s future in the AI economy. Energy and water use require proper oversight, but data centre operators are already investing in renewable energy, greater efficiency and recycled water because these measures increasingly make commercial as well as environmental sense. Regulation can strengthen and standardise those practices. The more fundamental questions are who will own and control Australia’s AI infrastructure, how resilient it will be, and where the economic value will be created.
The strategy largely treats AI as another technology requiring regulation. Yet AI is rapidly becoming something much bigger. Alongside telecommunications, cloud computing and digital identity, it is evolving into part of Australia’s critical digital infrastructure.
Once viewed through that lens, the policy debate changes fundamentally.
The important questions are no longer simply how AI should be regulated. They become: Who owns the infrastructure? Who controls access? How resilient is it? And how much long-term value will Australia actually create?
These are the strategic issues largely missing from the Government’s announcement.
AI is becoming critical infrastructure
Artificial intelligence is no longer just another software application.
It is becoming embedded in healthcare, education, finance, emergency management, defence, transport and government administration. Increasingly, AI will influence decisions that affect the everyday operation of the economy and society.
Governments already recognise that sectors such as telecommunications, electricity and banking require long-term infrastructure planning because society depends on them. Yet policy often remains reactive. As I criticised the response to the recent Telstra outage focused largely on the immediate operational failure rather than the broader challenge of building long-term national resilience. AI is now moving into exactly the same category. It should be treated not simply as another technology requiring regulation, but as critical infrastructure requiring long-term planning for sovereignty, resilience and national capability.
That requires a shift in thinking—from regulating technology to planning critical infrastructure.
Sovereignty is the first challenge
As I discussed in my article from only a week ago, for years Australia argued that Chinese telecommunications suppliers represented a national security risk because they ultimately operated under Chinese law. That principle has now become much broader.
The issue is no longer whether technology comes from China or the United States.
It is about jurisdiction.
Every technology company ultimately operates under the laws of its home country. I mentioned the recent events involving Microsoft’s obligations under the U.S. CLOUD Act, restrictions affecting AI services and the geopolitical debates surrounding Starlink all illustrate the same reality. These are not fundamentally American or Chinese issues—they are jurisdictional issues.
Australia’s AI future will depend heavily on companies such as Microsoft, Google, Amazon, OpenAI, Anthropic and NVIDIA. Their innovation is extraordinary, but Australia’s growing dependence on platforms beyond its own jurisdiction deserves far more strategic attention than the Government’s policy currently gives it.
Resilience matters as much as intelligence
The recent Telstra outage demonstrated that modern societies depend on resilient digital infrastructure.
The immediate cause was technical. The bigger lesson was whether essential services could continue operating when part of the system failed.
AI raises exactly the same challenge.
As governments and businesses increasingly depend on AI, resilience becomes just as important as capability. Australia needs to think beyond system reliability to three broader forms of resilience: technical resilience, cyber resilience and geopolitical resilience.
The above-mentioned discussion about digital sovereignty highlighted how foreign-controlled digital services may be affected by sanctions, export controls, national security directives or commercial decisions taken overseas. AI systems are exposed to the same risks.
The lesson from telecommunications is that resilience cannot be built after the event. AI should therefore be approached as critical infrastructure from the outset, with a long-term strategy based on diversity of suppliers, interoperability, contingency planning and reduced dependence on any single foreign jurisdiction.
The bigger economic question
Much of the Government’s announcement focuses on protecting Australian copyright and managing the environmental impact of AI data centres.
Both are worthwhile objectives.
However, they are regulatory issues rather than strategic ones.
The larger economic question is whether Australia will become a creator of AI value or simply a location where overseas companies build data centres.
Australia has abundant renewable energy, political stability and attractive locations for hyperscale infrastructure. These are genuine competitive advantages. But history suggests we should ask a familiar question.
Will Australia once again provide the raw inputs while others own the intellectual property, the software, the platforms and the long-term profits?
Hosting AI infrastructure is valuable.
Owning AI capability is considerably more valuable.
That means investing not only in data centres but also in Australian research, commercialisation, sovereign capability and trusted international partnerships.
The next step
The Government has made an important start.
It has recognised that AI deserves national attention and that appropriate regulation is necessary.
The next step is more ambitious.
Australia needs an AI infrastructure strategy.
Such a strategy would focus not only on how AI is regulated, but on who owns critical capabilities, how resilient they are, how strategic dependencies are managed and how Australia captures more of the value created by the AI revolution.
Energy efficiency, water consumption and copyright protection all matter. But they will not determine Australia’s long-term position in the global AI economy.
The defining issues are sovereignty, resilience and value creation.
If Australia gets those right, AI can become one of our greatest economic opportunities.
If we do not, we may simply become the place where somebody else’s AI infrastructure is built.
Paul Budde