Beyond Defence: Lessons in innovation from across industry

I have the privilege of closing today’s conference, which means two things. First, almost everything sensible has probably already been said.

So rather than give you another account of Defence’s priorities or another tour of emerging technologies, I want to use this time to ask a slightly different question.

What should industry be thinking about now in order to position itself for the future?

It is a question worth asking because there is a great deal changing around us.

The strategic environment is becoming less predictable. Technology is moving faster. The assumptions that shaped global supply chains over several decades are being challenged. And Defence itself is changing the way it develops capability and works with industry.

The 2026 National Defence Strategy has set a clear strategic direction. The new Defence Industry Development Strategy (DIDS) gives more specific signals about where Australia needs to build industrial capability and capacity. Across Defence, there is greater emphasis on experimentation, continuous development and delivery, minimum viable capability and earlier engagement with industry.

VIEW ALL

There is more to do.

I know many people in this room have strong views about the experience of working with Defence, and some of those criticisms are justified.

We need to reduce unnecessary complexity, engage more consistently, make decisions faster and continue becoming a better-informed and more capable customer.

The new DIDS is explicit about that direction of travel. But today, I want to turn the question around.

Because industry is not simply waiting for government to create opportunities. Companies make consequential choices long before an opportunity appears: where to invest, which expertise to build and retain, which risks to carry, and which partnerships will make them stronger.

Those choices shape whether you are ready when the opportunity arrives.

And when I look at the environment ahead of us, three qualities keep recurring: resilience, adaptability, and a willingness to share risk.

These are not uniquely Defence ideas. We can see them playing out across energy, advanced manufacturing, technology, logistics and other complex industries.

Defence is, however, an unusually demanding environment in which to apply them.

Our systems are complex and highly integrated. They operate in extreme environments. The consequences of failure can be profound. Technologies may need to remain relevant for decades while the threat they are responding to changes continuously.

That makes Defence an unforgiving test of an industrial model. But it can also make companies stronger.

For several decades, much of the global economy became extraordinarily good at optimising for efficiency. We built lean inventories, highly concentrated supply chains and just-in-time delivery systems that reduced costs and increased speed.

In stable conditions, that logic created enormous value.

But the pandemic exposed something we should probably have understood already: a system optimised around one set of assumptions can become remarkably brittle when those assumptions change.

Since then, we have seen conflict disrupt access to energy, fertiliser and critical inputs. Geopolitical competition is reshaping technology supply chains. Governments and companies around the world are reassessing dependencies that, a decade ago, barely appeared on a risk register.

I am not arguing against efficiency.

Defence has a profound obligation to use public resources wisely, just as every company in this room has to remain competitive.

But the lowest-cost supply chain is not necessarily the most valuable if it fails precisely when you need it most. And resilience is much richer than redundancy.

It is not simply about holding more stock in a warehouse. It is about understanding your dependencies and knowing which capabilities you cannot rapidly replace. It is about maintaining access to deep expertise, building trusted relationships before you need them, and designing systems with enough flexibility to respond when your assumptions turn out to be wrong.

This is the logic behind the greater emphasis on self-reliance in the 2026 Defence Industry Development Strategy. Self-reliance does not mean self-sufficiency.

Australia will not – and should not – try to make everything ourselves. We are deeply integrated with our allies and partners. AUKUS, Five Eyes science and technology collaboration, and our growing focus on co-development, co-production and co-sustainment all recognise that our collective industrial capacity is stronger than anything we could create alone.

But interdependence works best when each party brings something of value.

That raises a question I think every Australian company seeking to work with Defence should be able to answer: What is the strength you bring that Defence does not already have?

It might be specialist technology or manufacturing expertise. It might be access to a fast moving commercial market, an ability to scale, or a business model that lets you do something we cannot easily do inside government.

Sometimes it is simply a way of looking at how to address a mission that challenges our assumptions.

The best industry partners do not replicate Defence. They make us stronger by bringing capability and expertise we do not possess ourselves.

And sometimes the capabilities that matter most are not the ones attracting the headlines.

For more than a decade, Defence Science and Technology Group (DSTG) has led work on next-generation battery technologies for Defence platforms and systems.

Batteries may sound mundane beside hypersonics or quantum technologies. Until you remember that an advanced platform is only useful if its energy systems work safely, reliably and under extreme conditions.

Our scientists have built the facilities and expertise needed to understand how advanced battery technologies perform, degrade and fail. We have worked with Australian manufacturers to reduce the risks of introducing those technologies into Defence capability and supported advanced lithium-ion technologies progressing towards industrial production in Australia.

The lesson is not really about batteries. It is that resilience often lives in the enabling capabilities everyone takes for granted – until they are no longer available.

For industry, understanding where those dependencies sit may be just as important as chasing the most visible technology opportunity. Of course, resilience alone is not enough.

A perfectly resilient organisation, exquisitely designed for yesterday’s environment is not much use if the world has moved on. And the world is moving.

Artificial intelligence, quantum technologies, autonomy, biotechnology and advanced manufacturing are not developing neatly, one at a time. They are converging, interacting and changing the environment in which organisations operate while those same organisations are still trying to understand them.

This is why I am increasingly sceptical of any innovation strategy that depends too heavily on predicting the future correctly.

We must scan the horizon. We must understand technology trajectories. We must make informed choices about where to invest.

But in an environment this dynamic, no organisation will correctly predict every breakthrough or every disruption. Advantage will increasingly come from the ability to learn and adapt faster when the world does something you did not predict. That requires a different way of thinking about innovation.

For a long time, we imagined innovation as a relatively orderly process. Research produced discoveries. Discoveries moved into development. Development produced something that could be procured. Procurement put it into service. There is nothing inherently wrong with those activities. The problem is imagining that they always occur in a straight line.

In complex systems, knowledge moves backwards and forwards. Operational experience changes our understanding of the problem. A manufacturing constraint changes the technical solution. A new technology creates an option that did not exist when a requirement was written. An experiment fails and, if we are paying attention, teaches us something important.

The DIDS describes a shift towards continuous development and delivery: smaller increments of capability, faster decision cycles, earlier testing in real-world environments, and minimum viable capability delivered sooner and improved over time. That is a change for Defence. It is also a call to industry to engage with us differently.

Defence must define the mission and the operational need. That is our responsibility. Our warfighters understand the environment in which they operate and the effects they need to achieve. But once the mission is clear, we need industry involved early in developing how that need might be addressed. Because you understand things we don’t.

You know how a technology can be engineered and whether it can be manufactured. You understand the cost and challenge of scaling. You see where commercial technologies are moving. You understand your supply chains and, very often, you know which assumptions will collapse when an elegant prototype meets the real world.

This is what we mean by mission-driven innovation. Clarity about the problem. Flexibility about the solution.

The Advanced Strategic Capabilities Accelerator is built around that principle.

One of the approaches we have been experimenting with is the Innovation and Asymmetry Wargame.

Last year, we brought operators, scientists, technologists and other experts together to explore emerging science and technology options against the operational challenges that matter most.

The purpose was not to select a winner in a technology beauty contest. It was to learn: to expose assumptions, see technologies in combination, identify where something apparently promising failed to create operational advantage and find opportunities we might otherwise have missed.

The insights from that work are now helping inform future investment decisions. There is a broader industrial lesson here.

Experimentation is not what you do after you have eliminated risk. Done properly, experimentation is how you understand risk early enough to make better choices. That is an important distinction.

I have spent much of my career in science organisations, and one of the privileges of that experience is seeing failure at close quarters.

Science progresses because we are prepared to discover that an idea is wrong.

The discipline is not avoiding being wrong. It is designing the experiment so that, when reality disagrees with you, you learn something valuable. The same principle applies to innovation.

The companies that thrive in a faster-moving environment will not necessarily be those that arrive with the most polished first answer. They may be those most capable of recognising that their first answer was wrong, extracting the lesson and adapting before everyone else.

That is what adaptability looks like in practice. But adapting early creates another challenge. It asks people to make decisions before every uncertainty has been removed. And that brings us to risk.

We use the word partnership a great deal in Defence. Sometimes too easily.

A contract does not automatically create a partnership. Nor does collaboration mean that everyone sits around a table until responsibility becomes impossible to locate.

Real partnership starts with recognising that the parties bring different strengths and have different responsibilities.

Defence understands the mission, the operational environment and the consequences of failure.

Our science organisations maintain deep specialist expertise, including in sensitive and classified domains, and provide the evidence needed to reduce technical and capability risk.

Industry knows how to engineer, integrate, manufacture and scale. Universities expand the boundaries of what is possible.

Our allies and partners extend the expertise, industrial capacity and technology available to us.

The power comes from combining those strengths around a shared goal. But doing that requires a more mature conversation about risk.

Defence cannot ask industry to invest against opaque priorities and uncertain demand.

That is why the clearer investment signals in the DIDS matter. The more granular Sovereign Defence Industrial Priorities matter. The work on alternative financing and a Defence Industry Investment Plan matters.

Defence has a responsibility to be clearer about where we need capability and where we are prepared to invest. But clarity from Defence does not remove every commercial risk. Nor should it.

There will not always be a guaranteed program at the end of an idea. Technologies will fail. Operational needs will evolve. Some approaches will prove impossible to integrate.

If our model requires government to absorb all uncertainty before industry acts, we will move too slowly.

Equally, if Defence transfers unreasonable risk to its suppliers – particularly small and medium-sized enterprises (SME) – we will hollow out the very industrial base we are trying to build. So perhaps we have been asking the wrong question.

The question is not simply: who carries the risk? It is: who is best placed to understand, reduce and carry each part of the risk?

I think that is a much more mature conversation and one we need to become much better at having.

It also changes what we mean when we talk about a win-win relationship between Defence and industry.

A win-win is not Defence lowering every barrier until a company gets a contract. Nor is it industry accepting every condition because Defence is the customer.

The real win comes when each party contributes strengths the other does not have, risks are held where they can best be understood and managed, and the result is a capability Australia could not otherwise have created. That is partnership.

And there is one capability that increasingly determines whether those partnerships succeed: integration.

If you look at the seven Sovereign Defence Industrial Priorities in the new DIDS, integration is everywhere: development and integration of autonomous systems; integration and enhancement of battlespace awareness and management systems; test, evaluation, certification and systems assurance.

Or look at Defence’s innovation, science and technology priorities: AI, quantum, undersea warfare, long-range fires and hypersonics, high-energy lasers and autonomous systems.

None of these technologies creates decisive capability in isolation.

A brilliant sensor that cannot integrate has limited value. An autonomous platform that cannot operate in tandem with crewed platforms has limited value. An AI tool that cannot be used in coalition operations has limited value.

In complex capability systems, solutions that cannot integrate are not solutions. This matters particularly for SMEs.

You may possess a genuinely world-leading technology, and Australia needs companies with that ambition and depth. But Defence rarely buys isolated technologies.

We build systems of systems. Those systems have interfaces, dependencies, security requirements and long sustainment tails. Increasingly, they also need to operate with our allies and partners.

So one of the most important capabilities a company can develop is the ability to play well with others.

I use that phrase deliberately. Because collaboration is sometimes discussed as if it were a soft virtue – something desirable but secondary to technical excellence.

In complex systems, it is an engineering and industrial capability.

Can you understand interfaces? Can you work with a prime, another SME, a Defence laboratory or a university whose incentives and ways of working differ from yours? Can you recognise when another organisation brings a strength you need?

And can you accept that your greatest value may sometimes come from being a small but indispensable part of a larger capability rather than owning the whole solution?

These questions matter beyond Defence.

In fact, the same forces are reshaping complex supply chains across the economy.

As products become more software-enabled, as AI is integrated into business processes, and as technologies converge, fewer organisations can hold every capability they need inside their own walls.

The capacity to integrate expertise becomes an advantage in its own right. That is also why dual-use technologies matter so much to Defence.

Many of our highest technology priorities are deeply rooted in commercial markets. AI is an obvious example, but the same is true of autonomy, quantum technologies, advanced manufacturing and many areas of sensing and communications.

Defence must make greater use of technologies developing in those markets.

But I want to challenge another assumption.

The logic does not run equally in both directions.

Commercial success does not automatically make a technology Defence-ready.

Defence systems face different environments, security requirements and assurance requirements. They must integrate with legacy systems and, increasingly, with coalition systems. And sometimes the consequence of an incorrect answer is measured in lives.

Working successfully with Defence can create disciplines, technologies and integration capabilities that support success in other markets. But a successful commercial product cannot simply be rebadged as a Defence capability.

For industry, understanding that distinction is part of becoming a strong Defence partner.

And nowhere are the questions of expertise, integration and judgement more important than in artificial intelligence.

AI is already changing how organisations operate. It is changing how software is developed, how information is synthesised and how data is processed.

Increasingly, it is changing how decisions are made.

On the battlefield, specialised AI systems are already being used for detection and tracking, autonomous systems, enhanced command and control and battle-damage assessment.

For military forces, using AI effectively is not optional.

A force unable to exploit it will cease to be operationally credible.

But as we race – appropriately – to understand and use these technologies, I think we need to be very careful about one assumption that sometimes accompanies the discussion.

Nearly a century ago, T.S. Eliot asked a question that feels unexpectedly contemporary: “Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information?”

Artificial intelligence gives that old question new urgency.

We now have tools capable of retrieving, combining and generating information at a speed and scale Eliot could scarcely have imagined. But information is not knowledge, and knowledge is not judgement.

The fact that an answer can be produced instantly does not tell us whether we have asked the right question. It does not tell us whether the data on which the answer rests is meaningful, whether an assumption that was valid yesterday remains valid in a different environment, or whether a result that looks entirely plausible is actually true.

The more capable AI becomes, the more dangerous it will be to confuse access to an answer with possession of expertise.

I want to be clear. I am not making an argument for against the use of AI. Quite the opposite.

If AI can do part of our work faster and better, we should use it. We should redesign the way our organisations operate so that AI amplifies human capability.

Scientists should not spend their time doing work a machine can do better. Neither should engineers, analysts, managers or CEOs.

But we also need to understand what deep expertise actually is. It is not simply the possession of information.

It is accumulated judgement. Tacit knowledge. The memory of what has been tried before. The ability to recognise the anomaly that matters, or to sense that a beautifully presented answer is inconsistent with how the physical world actually behaves.

And, critically, it is the capacity to take responsibility for a decision when the evidence is incomplete.

This matters enormously for science organisations like DSTG. Our future role is not to compete with AI in producing answers.

It is to maintain and continually renew the depth of specialist expertise Australia needs to know which answers can be trusted.

That changes how I think about the stewardship of scientific capability.

For decades, we have often described the value of a science organisation by pointing to what it produces: discoveries, technologies, patents, reports or capabilities.

Those things matter.

But perhaps one of the most important things a science organisation does is sustain communities of deep expertise over time.

People learn from each other. They inherit knowledge that was never fully captured in a report. They develop experimental facilities and methods. They remember why an apparently obvious solution was rejected 20 years earlier – and know when a change in technology means it is worth trying again.

AI can reshape how those communities work. It can make them more productive.

But if we mistake information retrieval for expertise and allow the underlying human capability to atrophy, we may discover too late that we have lost something very difficult to rebuild.

I suspect the same is true of many organisations represented in this room. Everyone will have access to powerful AI. Access alone will not distinguish you.

Your competitive advantage may increasingly depend on the expertise you have built around it: the people who understand your technology, your manufacturing processes, your customers, your failure modes and your systems.

AI can accelerate learning. But only an organisation that retains the capacity for judgement knows what it has learned.

So, where does all of this leave an Australian company thinking about its future relationship with Defence?

I don’t think there is a magic formula.

But from my vantage point – working with operators, scientists and engineers, industry leaders, universities and international partners – I do see a pattern.

The strongest partners bring us something we do not already have. They take the time to understand the mission before falling in love with a solution, but once that mission is clear, they bring creativity and expertise to how it might be addressed.

They are prepared to learn and adapt because they understand that the first prototype is rarely the capability and the first answer is rarely the final answer.

They think about integration early. They understand that their technology has to exist in a system, a supply chain and increasingly an alliance environment.

They are willing to work with others, including organisations that may bring competing or complementary strengths.

And they are prepared to share risk – not recklessly, and not by accepting risk that properly belongs with Defence, but by recognising that genuine innovation cannot begin only once every uncertainty has been removed.

Those are demanding expectations of industry.

Defence must meet equally demanding expectations of its own.

We need to provide clearer priorities, engage better and make decisions faster. We need more intelligent procurement and a deeper understanding of the capability and capacity of our industrial base.

And we need to work continuously with industry rather than appearing at either end of a transactional procurement process.

That is the direction set by the new Defence Industry Development Strategy.

The opportunity now is to turn policy intent into a different way of working. I want to finish with a familiar Australian story. But perhaps a less familiar lesson from it.

The Jindalee Operational Radar Network (JORN) is rightly recognised as one of Australia’s great science and technology achievements.

More than 50 years of science, engineering and innovation created a capability many thought impossible.

And last month, through our partnership with Canada, the technology that underpins JORN became the basis of Australia’s largest-ever defence export.

It is tempting to tell that story as proof that Australians can invent extraordinary technology. We can.

But I think that misses the more important lesson.

The people who began the work that became JORN could not have specified its final destination.

They could not have predicted the technologies that would emerge over the following five decades. The operational environment changed. The science itself kept revealing new problems.

And they certainly could not have known that one day, the very different ionospheric physics of the Arctic region would create a new scientific challenge to tackle with one of our closest partners.

What Australia sustained was not simply a technology.

We sustained the capacity to keep solving the problem.

That capacity lived in deep scientific expertise, industry capability, operational understanding and international partnerships. It lived in generations of people prepared to build on the work of those before them and to adapt as the problem changed.

That is resilience. It is adaptability.

And it required government, science and industry to carry different risks over a very long period of time.

The lesson from JORN is to ensure we are building the conditions from which the achievements we cannot yet imagine will emerge.

Government has choices to make. Defence has choices to make. And industry has choices to make.

You are not simply supplicants waiting for the next grant, contract or program. You are part of Australia’s national capability, and the investments you make now, the expertise you retain, the partnerships you build and your willingness to adapt and share risk will help determine the choices Australia has in the future.

We cannot predict every disruption. We cannot know which technology will next change the strategic balance.

But we can decide what kind of innovation and industrial ecosystem we build: whether it is resilient enough to withstand shocks, adaptable enough to learn as the world changes, and connected enough to bring together the different strengths of Defence, industry, science and our international partners.

The future will ask us questions we cannot yet imagine.

Our responsibility is to make sure Australia keeps the capacity to find the answers.

Thank you.

Professor Tanya Monro AC presented this address at the inaugural Australian Defence Industry Accelerator Summit at the National Convention Centre in Canberra on Thursday, 16 July 2026.