When it comes to innovation, Australia has a strategy problem as much as a delivery problem. The first is about purpose. The second is about execution. I recently argued that Australia’s innovation system lacks the strategic coherence and delivery posture required to match the pace of threat in a contested Indo-Pacific. But solving that problem starts with one crucial question: how do we get our best ideas across the finish line?

Translation is a single but foundational shortcoming within any innovation lifecycle. Even when Australia produces world-class research and breakthrough technologies, we routinely fail to translate those into sovereign capability. It’s not a discovery gap. It’s a deployment one. Australia lacks the institutional, financial and operational infrastructure needed to move from invention to impact. The result is an innovation engine that spins but rarely delivers.

We are a nation of inventors but have so far lacked a consistent path to impact. Wireless LAN, developed by CSIRO in the 1990s, was an early precursor to Wi-Fi which became a global standard. But the commercial value of the initial research flowed largely offshore. The cochlear implant, Google Maps’ core algorithm, polymer banknotes and the black box flight recorder all began in Australia. Few were scaled or retained here. In today’s strategic environment, where cyber tools, sensing and software define deterrence this failure is more than economic. It’s a national security and a national safety risk.

This isn’t due to a lack of talent—Australia’s researchers, engineers and founders are globally competitive—nor a lack of ideas. From quantum sensing to AI-enabled security, our innovation base regularly produces capabilities aligned with AUKUS Pillar Two. But the system is not built to carry those ideas from prototype to platform. The so-called valley of death is a design flaw.

Promising technologies lose momentum when public funding ends and private capital won’t assume the risk. Others are acquired offshore before they can be deployed here. Procurement doesn’t catch them, commercial capital doesn’t back them and no agency is charged with bridging the gap.

The challenge is amplified by Australia’s persistent tendency to adopt technologies rather than create them, and by structural underinvestment in both research and commercialisation. Government spending on research and development as a percentage of GDP has declined from 0.56 percent in 2007–08 to just 0.52 percent in 2024–25. These are not the metrics of a country building a resilient, sovereign innovation base.

This dynamic is especially troubling in areas directly relevant to national safety and resilience. Secure identity platforms, counter-drone systems, AI-driven threat detection and infrastructure protection are just a few examples where capability shortfalls exist—and where domestic innovation could deliver. Yet there are no mechanisms to scale these technologies from the lab into the hands of the agencies that need them.

While Defence has made some headway with the creation of the Advanced Strategic Capabilities Accelerator (ASCA), other high-risk, high-consequence agencies, such as Home Affairs, lack similar tools. Without a clear expression of demand, there is no signal to guide founders, funders, or researchers. And without a path to procurement, sovereign technologies will continue to stall, stagnate, or be lost.

Other countries do this better. The United States funds early-stage, mission-aligned innovation through the Small Business Innovation Research program. Britain runs challenge-based cycles through the Defence and Security Accelerator. Israel embeds translation into its national innovation architecture, backing early-stage technology through military-academic collaboration, dedicated venture structures and procurement pathways shaped by operational need. All three treat translation as a core capability—not an afterthought—and accept that some risk is necessary to maintain a technological edge.

Australia has begun to explore similar models. The National Reconstruction Fund and ASCA represent progress. But they are still fragmented and narrowly focused. ASCA remains confined to Defence. The National Reconstruction Fund is structured around sectors, not missions. Neither provides a dedicated mechanism for Home Affairs or civil-security agencies to act as launch customers for trusted, sovereign technology.

Meanwhile, Australian startups continue to build promising tools in quantum sensing, edge computing, identity assurance, cybercrime analytics and secure supply chain visibility. These align directly with AUKUS Pillar Two, but without scale, they exit. And when they exit, so does our control.

Translation is not just a policy challenge; it’s a sovereignty one. If we cannot build, deploy, or trust the systems we need, we remain economically, operationally and strategically dependent on others.

To fix this, we need clear demand signals, simplified procurement for early-stage technologies, and public-private pathways for mission-specific prototyping. Translation must be seen not as a transition phase, but as a core capability of the innovation system

Australia does not lack good ideas; it lacks the infrastructure to carry them across the finish line. Until that changes, the most critical capabilities will continue to be the hardest to field.

If translation is the valley of death, then capital—how it is allocated, de-risked and aligned—is the terrain we must learn to navigate next.