Media
|
Speeches
October 16, 2026
Thank you. It is a privilege to be at Consilium once again.
For 50 years the Centre for Independent Studies has been a respected and influential voice for classical liberal values.
Your research has shaped policy debate and contributed to a more prosperous and free Australia.
I’m proud to have been associated with the CIS for almost twenty years since I first completed the Liberty and Society Program as an undergraduate student.
Greg Lindsay is one of Australia’s great intellectual entrepreneurs and you have all built a lasting institution committed to the power of ideas.
Let me start with a question.
If a conflict over the Strait of Hormuz, on the other side of the world can wreak such havoc to Australia’s supply chains, living standards and economy, what would a closure of the Malacca, Lombok and Sunda Straits do to Australia?
Closure of these maritime chokepoints in our region would make the Hormuz crisis look like a walk in the park.
98 per cent of our trade is seaborn. The vast majority of it transits through these waters, and carries with it products essential to our economy and way of life – like almost all our refined fuels, not to mention our exports which pay for them.
It would be an economic and security disruption far more serious than COVID.
And in today’s disrupted world, its no longer a far-fetched scenario.
The past eight months have seen a US–Israeli war with Iran that killed its Supreme Leader, a fifth year of all-out war in Ukraine, and a nuclear-capable missile fired into the South Pacific by the People’s Liberation Army.
Just three weeks ago, People’s Republic of China vessels aggressively blocked a Philippine resupply mission in the South China Sea, while military exercises around Taiwan increase in scale and frequency.
None of these, on its own, is the story.
The story is that they have been happening at the same time, and that they are all interlinked.
In January, Canada’s Prime Minister Mark Carney told Davos that “we are in the midst of a rupture, not a transition.”
I'd put it slightly differently. The rules-based order isn't dead. But it is on life support.
The order my generation grew up with in the 90s and 2000s rested on three things: American primacy, high barriers to aggression, and widespread belief in an open global market that made war costly and trade rewarding.
None of these would have worked without the others: global trade depended on a US-led provision of global public goods – in particular, secure shipping lanes.
And now, all three are weakening at once.
So I would not say we are living through a collapse of the global order, but of its fragmentation, in several theatres at once, accelerated by technologies that make warfare cheaper, and by a retreat from the open markets that made us prosperous.
I want to take you through three topics: where the fires are burning, what is fueling them, and some suggestions for what we must do to successfully navigate it.
Ukraine
First, Ukraine.
After more than four and a half years, Russia's full-scale invasion, which many predicted would be over in days if not weeks, continues to grind on.
In the four weeks to late September, Russia gained a net 37 square miles, after losing ground the month before.
These gains may have again been wiped out by Ukraine’s most recent, secret offensive.
But the back and forth on the map disguises several concurrent revolutions in warfare.
Ukraine's long-range drones and missiles have taken around a quarter of Russia's oil refining capacity offline.
Ukraine is a test bed for modern tactics and technologies.
Every new technology is being proven, or failing, there first.
Estimates vary, but between 70 and 90 per cent of Russian casualties on the front line are believed to have been inflicted by drones.
Hundreds of thousands have been killed, driven in part by the total disregard of Putin’s regime for human life – including that of his own citizens.
If you want a shocking insight into the horror of modern war, listen to the School of War podcast by Aaron Maclean from July, where he interviews a former US special forces operator who has spent four years fighting in Ukraine.
In it he details some of Russia’s brutal tactics – including marching soldiers into the no-go zone on the front line to certain death, just to identify where Ukrainian drones are so they can be destroyed.
In this grim game of chess, Russia is treating human beings as mere pawns, happily sacrificed to identify what they clearly believe are more militarily valuable machines.
The Middle East
Second, the Middle East.
Despite the stunning tempo of US and Israeli operations, which have delivered record amounts of ordinance to thousands of targets in just the first few weeks of the war, the Islamic Republic regime remains defiant.
They have so far successfully held off the world’s most powerful military with hundreds of missiles and thousands of drones fired across the Gulf, at American bases, at oil infrastructure, at shipping.
And despite the losses it has sustained, Iran has demonstrated it can do something that matters a great deal to us.
It can close the Strait of Hormuz with the mere threat of violence, demonstrating just how fragile open trade is when an actor, no matter how weakened, is determined to end it.
Today there is a US naval blockade, a tit-for-tat tanker war, and in August alone Iran struck at least 13 commercial ships.
Iran's Houthi proxies are seeking to do something similar in the Red Sea.
It is a remarkable demonstration of the power of asymmetry in modern warfare, which I will return to shortly.
Indo-Pacific
Third, and most important for us, the Indo-Pacific.
On the 24th of September, PRC coast guard and naval vessels obstructed a Philippine resupply mission to Second Thomas Shoal, the first such incident in two years.
Three days later, Beijing staged joint naval and air drills around Scarborough Shoal.
Around Taiwan, last December’s PLA exercises rehearsed blockading ports, with 130 aircraft around the island in a single day.
And in July, the People’s Liberation Army-Navy test-fired a nuclear-capable missile into the South Pacific, close to island states we consider family.
To sustain our liquid fuel needs we have had to look to Japan, South Korea, Singapore, and the PRC itself, highlighting how dependent we have become on south-east Asian shipping lanes surrounding the South China Sea.
Xi Jinping has conducted a near-total purge of the Communist Party’s Central Military Commission.
Five of the PRC’s most senior generals – all of whom were first appointed to the Commission by Xi – have been arrested and removed from this critical military decision making body.
Just two remain – Xi himself and Zhang Shengmin. Talk about a stressful work environment.
Analysing CCP factional politics is notoriously fraught. But we can be reasonably confident based on official statements from the Party that this is not simply an issue of corruption, with references made to loyalty to the party.
It seems clear, especially following earlier purges to the PLA Rocketforce and associated state-owned enterprises, that Xi is seeking to build a more loyal and compliant PLA, willing to faithfully follow his directions in all future contingencies – including the possible annexation of Taiwan by force.
Lessons to learn
These three theatres are not separate.
They are interconnected. We must learn the lessons from the Middle East and Ukraine and apply them to our own region, now, while we still have time.
As General David Petraeus and the historian Andrew Roberts argue in their superb book Conflict: ‘A military should always strive to be a ‘learning organisation’, especially in war.’
The new unfolding trends driving these conflicts are threefold: drones, missiles, and AI.
Drones
Starting with drones.
On the offensive side, Ukraine went from producing a few thousand FPV drones a year in 2022 to around three million last year, with capacity for more than eight million.
Russia fired more than 54,000 (Iranian) Shahed-type drones at Ukraine in 2025 alone.
Ukraine’s Operation Spiderweb used 117 cheap drones, smuggled into Russia on trucks, which damaged or destroyed more than 40 Russian aircraft worth over seven billion US dollars.
A Patriot missile designed to intercept drones costs more than 3 million US dollars. A Ukrainian interceptor drone costs around 2,500 US Dollars.
As former Major General Mick Ryan observes, Russia is moving beyond the slow, propeller-driven Shahed to jet-powered drones like the Geran 4 and Geran 5 that fly higher and faster.
They are forcing Ukraine to scramble fighter jets and fire expensive Patriot interceptors.
Ryan describes one recent Ukrainian operation as a ‘large, real-time experiment and proving ground.’
And he poses the question that will decide this European winter: can Ukraine produce faster than Russia can?
It is a question every middle power, including ours, should be asking itself too.
Drones that are relatively modest in capability have become an essential tool for Ukrainian and Russian forces.
Ukraine did not achieve this with a five-year central plan.
More than 500 private firms compete to supply its military.
Design cycles run in weeks, not years, and Ukraine now produces the most capable military drones in the world.
Missiles
Second, missiles.
When Iran fired its barrages across the Gulf, states with the best American air defences watched their interceptor stocks drain.
And now the order backlog to replenish those stocks stretches for years into the future.
Lockheed Martin built roughly 600 of its most advanced Patriot interceptors last year.
Ukraine used about 700 in just four winter months.
The Iran War has made this an Indo-Pacific problem.
Analysts at CSIS estimate the United States fired close to half its Patriot stocks in the first seven weeks.
Mark Cancian, one of the report's authors, warned that this created “a window of increased vulnerability in the western Pacific.”
Restocking will take years.
In the next war, the question won’t be who has the best missiles.
It will be who still has missiles after the first week, and who can produce them faster than they need to use them.
Right now, we can’t.
Artificial Intelligence
Third, AI.
AI is transforming modern warfare.
In the first 38 days of the Iran war, the United States struck more than 13,000 targets, using AI systems that fuse intelligence and generate strike options in minutes.
CENTCOM’s commander said these tools help commanders “make smarter decisions faster than the enemy can react.”
Mick Ryan observes that with algorithmic warfare, hypersonics and autonomy, “war is speeding up and actions will be measured in micro-seconds.”
Even cells in Yemen are making use of Claude to code their guided weapons and radically close technology gaps.
The Israeli Harpy drone searches areas for enemy radars which it can then destroy without requesting human permission, something which China has already reverse-engineered.
I recently visited the Melbourne facility of an American defence contractor which produces independent, AI-piloted fighters and drones which do not rely on communication with the ground.
As Petraeus and Roberts argue,
“the advantages robots enjoy on the battlefield are immense...they can have far greater resilience, require no sleep, can cross any terrain and can take part in suicide missions that rest on the conscience of neither themselves nor their operators. The idea that they will not be employed by countries out of moral considerations is ludicrous; rather than presenting a moral Rubicon, many nations consider robotics a moral boon.”
Soon, every modern military platform will be AI enabled in some way.
For a middle power like Australia, access to this technology is even more important.
We can never compete in raw numbers.
But a nation with a smaller population can significantly enhance its lethality with autonomous systems to deter a much larger potential adversary.
In a major-power conflict, the states who have access to the most advanced AI models will have measurable battlefield advantages over their adversaries.
The risk is not only on distant battlefields.
Last month, an AI agent went somewhere it was not sent, inside an Australian government system.
While no harm occurred this time, we should treat this as a free warning.
Next time it could be Defence, the ATO or our electricity grid.
Frankly, I would be amazed if the Services Australia Medicare Statistics Portal incident is actually the first or only time an AI agent has breached our systems.
Our intelligence agencies have publicly assessed that PRC cyber actors including the PLA’s Cyberspace Force and the Ministry of State Security’s cyber teams have for years been identifying and seeking to exploit vulnerabilities in our critical networks.
If they are not using China’s best AI models to do this today they are guilty of professional misconduct and will be in line for Xi Jinping’s next purge.
The simple truth is that AI can identify and weaponise network vulnerabilities faster and more effectively than humans ever could.
Our only defence against AI-enhanced cyber attacks is even better AI-enhanced cyber defence.
If we, the West, do not continue to advance, the PRC will, and they will be able to apply AI both on the battlefield, as well as to mine our personal data, our intellectual property, and exploit every vulnerability.
That is why access to the world's most capable models is a national security imperative for Australia.
So what actions must we take? Let me offer five brief suggestions.
Tell the truth
First, we must tell Australians the truth.
If Australians knew how likely conflict in our region is, and how unprepared we are, they would be marching in the streets for higher Defence spending.
Instead, they are being lulled into a false sense of security.
The public wants candour.
ANU research found that 53 per cent of Australians believe government shares too little about security threats.
Spend real money
Second, we must spend at least three per cent of GDP on defence, following the recommendations of the government’s own hand-picked reviewers, Sir Angus Houston and Professor Peter Dean.
And we must do so with urgency, heeding the warnings of our best informed experts that we have no time to spare.
Build more self-reliance
Third, we must build self-reliance in partnership with allies and other like-minded nations.
The United States remains indispensable.
Every one of our most important partnerships, from Five Eyes to AUKUS to the Quad, is anchored by Washington.
But the alliance can no longer be our only strategy.
That means sovereign production of missiles and munitions, so we are not at the back of a very long American queue for resupply.
It means integrated air and missile defence for our northern bases and our critical infrastructure.
It means embracing the defensive asymmetry Ukraine has shown us, with cutting-edge and cost-effective drones.
AUKUS Pillar II – pillar of delivery
Fourth, we must make AUKUS Pillar II – the technology pillar - deliver, and let private capital in.
The first major Pillar II project, common payloads for uncrewed undersea vehicles, from 2027, is welcome but too little after five years.
To begin with, we must cut remaining regulatory barriers, as roughly 20% of intra-AUKUS defence trade still falls outside licensing exemptions.
Securing sovereign AI
Finally – we must secure sovereign Australian access to the most advanced frontier AI models from the United States.
We have a brief and rapidly closing window to do so.
In a year this opportunity will be gone.
American frontier labs need a second home of training.
It must be a US ally, a rule of law democracy, and with the room to build and the natural resources to power their most advanced chips in modern data centres.
It also needs a regulatory environment compatible with its business model.
That’s a challenging policy problem but not an unsolvable one with goodwill and parties willing to put the national interest first.
Realistically only a handful of countries fit this bill.
The one that does will unlock potentially hundreds of billions of investment, and guaranteed access to the world’s best models because they will be trained onshore.
They will also get a seat at the table to influence global AI regulation that’s massively disproportionate to their size and global stature.
The alternative is not that we can halt AI development.
It is that we instead choose to be a bystander with no agency, no access and no influence.
In a dangerous world that’s a reckless choice.
Our world is strategically disordered and technologically dynamic. To navigate the first challenge, we must master the second.
As Professor Rory Medcalf has written, “a path can be charted between conflict and capitulation.”
If we make the right choices, we can prevent conflict spreading to our region with credible deterrence.
But we don’t have any time to waste.
Thank you.
ENDS