Learning From Ukraine: Semi-Autonomy Can Deliver Combat Mass

Generating operational mass in the Indo-Pacific requires new approaches to projecting, sustaining and employing military power across vast distances and contested environments. This must be achieved despite constraints in workforce and industrial capacity that can undermine sustainment in high-intensity warfare. A possible solution to this challenge is to embrace rapidly evolving, semi-autonomous uncrewed systems, while adapting defence industry to exploit rapid, continuous modernisation.

At the 2026 ASPI Defence Conference in June, a panel explored the role of uncrewed semi-autonomous systems. The discussion focused on developing operational mass to enable the ADF to better project, sustain and employ military power across the Indo-Pacific.

In the past four years, what we’ve been able to achieve is a unique ecosystem where we have soldiers, where we have startups, where we have volunteers and government agencies all fused together, working on products where you can develop a product from prototype to actual use on the battlefield in a matter of weeks.

He noted that even long-range missiles could be produced in around 18 months, and he argued that advantage lay in an ‘innovation loop from the battlefield information to manufacturing, R&D and then to basically providing it back to the battlefield.’ The key is ‘not how you innovate but how quickly you adapt to change, because three months later it becomes obsolete.’ He also raised electromagnetic warfare and uncrewed systems as important areas for collaboration between Ukraine and Australia.

For Australia, where capability development and acquisition processes often last years or even decades, applying this rapid innovation and adaptation model would be challenging. But it’s one that Australia should embrace to keep pace with China’s growing military power. Air Vice-Marshal Nicholas Hogan, head of air force capability, brought up industry’s role during the panel:

It takes a bit of time to get industry right, and then once industry’s working and then it starts moving forward … then they’re able to pump out at mass and then decrease the per-unit cost.

Still, it’s not simply a case of industry making the same item quickly and in large numbers. As Mark Palmer, vice president at Shield AI Australia, said:

What we’ve learned is that you have to be able to evolve. You have to be able to look at weaknesses, look at what’s being exploited and you have to evolve. It’s a continual game against each other on the tactics, and it’s almost always being won by the one that evolves quicker.

Palmer made the further point that even semi-autonomous systems were not immune: ‘if it’s autonomy that doesn’t evolve, doesn’t change, it can actually become a weakness.’

Real evolution often occurs in wartime, in a different operational environment than during development and testing. Modern warfare takes place in a highly contested environment, often with degraded communications and denial of critical capabilities, such as GPS.

So, for Australia the lessons from Ukraine are vital to developing future semi-autonomous systems with a future highly denied environment in mind. Rather than a human in the loop remotely operating a drone, be it an MQ-28 Ghost Bat or an MQ-4C Triton, the ADF may need to shift to people being on the loop – only overseeing what the machines do. This will allow it to rely more heavily on AI and autonomy to make critical mission decisions in a complex and fast-moving military operational environment.

Drawing on Shield AI’s experience in Ukraine, Palmer noted that this approach ‘allowed us to put some of the decision making into the front line’. For these capabilities to reach their full operational potential, it will be essential to devolve that decision-making on how best to use autonomous systems down the chain of command. That will be a major military and political challenge for the ADF, given Australia’s need to comply with obligations under international humanitarian law and given a political need for oversight of rapidly moving military operations.

Hogan highlighted this issue when discussing accountability and AI as potential barriers to generating operational mass. He argued that ‘the barrier from our side of the house is looking at the way we culturally adapt, because data isn’t the problem. It’s the software, and the amount of autonomy that we allow us to take on with the social licence that may be granted to us by the Australian public.’ Getting past this cultural barrier was critical for industry to adapt and deliver capabilities quickly, he said.

For Australia – as for Ukraine – time is short. Myroshnychenko noted that, in relation to the war in Ukraine, ‘there’s so much at stake for Australia’. And beyond it was the larger struggle with Russia and China. The central challenge, he argued, was not necessarily about technology but rather about ‘the leadership of the country and civil society and how well aware they are about the threats’.

Responding to this challenge will require governments to embrace innovation, recognise the rapidly deteriorating global security environment and adapt to how military technology is changing the character of warfare. Ukraine has been forced to make that shift, and the question now is whether Australia will do so before circumstances demand it.

Myroshnychenko concluded, ‘it’s very difficult to achieve what we want to achieve in a peaceful environment where there’s no visible threat or it’s never actually in a public domain.’ The starting point for exploiting semi-autonomous systems within the ADF must therefore be a frank conversation with the Australian public about the prospect of future conflict and how we must adapt to that emerging reality.

Malcolm Davis is a senior analyst at ASPI.

Image of Vasyl Myroshnychenko speaking at ASPI’s 2026 Defence Conference – ASPI.

AI contributed no ideas to this article – Malcolm Davis.

This article appeared originally at The Strategist (ASPI).