- Posted on 28 Sep 2026
- 3-minute read
More AI alone is not the solution to prosperity but how and when to adopt the technology is key to building lasting growth.
This is an adaptation of a speech delivered at the launch of ‘Turning AI into Productivity: The role of Innovation and Industrial Ecosystems’; a UTS research project supported by the Google Foundation and led by Emeritus Professor Roy Green AM with Dr John H Howard.
Key takeaways
- AI will only lift productivity if it is embedded in real industries, not simply adopted as standalone tools.
- Australia needs stronger innovation ecosystems so AI capability translates into sovereign industrial strength and practical adoption at scale.
- Successful AI adoption is an organisational and policy challenge, not just a technical one.
There is no shortage of talk about AI, what it can do, how quickly it is moving, whether we are keeping up, and, inevitably, when it might take over.
Turning AI into productivity, a report released this week, creates space for a more useful conversation: not simply about adopting AI, but about what it will take for AI to lift productivity in the real economy and build lasting prosperity.
Australia will not become more prosperous simply by using more AI. We will become more prosperous if we use AI to build industries, capabilities and jobs here in Australia.
Three points stand out.
First, we need to look beyond generative AI. The bigger opportunity is industrial AI: AI embedded in products, production systems and services, in manufacturing, energy, transport, logistics, health and agriculture.
Think about an Australian manufacturer using AI to detect faults before equipment fails, reduce waste, improve energy use and help skilled workers make better decisions. The value is not the algorithm on its own. It comes from combining the technology with engineering knowledge, good data, changed workflows and people who know how the operation actually works.
Second, connection matters. Good research, capable firms and talented people are not enough if they remain disconnected. We need ecosystems, and institutions, that connect research with industry, demonstration, procurement and adoption. Proximity helps. But relationships, shared purpose and the capacity to turn ideas into use matter more.
This is a natural agenda for UTS. We are a technology university with a strong record of working across disciplines and with industry. We do not want simply to comment on AI adoption. We want to help Australia do it well, to turn research and technological capability into stronger industries, better jobs and practical impact.
That is also why we are developing an Industrial AI Cooperative Research Centre bid with industry and research partners. The proposed CRC is focused on embedding AI in physical production systems and building Australia’s advanced manufacturing and sovereign capability. It is a practical way to bring research, industry demand, skills and adoption together.
Third, adoption is an organisational challenge, not just a technical one. Leaders need to know where AI will genuinely add value, how work needs to change, how people will be involved, and how risk will be managed.
Buying a tool is easy. Changing an organisation is hard.
That point is especially important for smaller firms. They need practical help: access to facilities, demonstrators, trusted advice, management capability and customers willing to be early adopters.
Leaders need to know where AI will genuinely add value, how work needs to change, how people will be involved, and how risk will be managed.
The issue is not whether Australia should engage with AI. Of course we should. The issue is where we can create real value, what capabilities we need, and how we move from isolated examples to adoption at scale.
That means joining up research, skills, infrastructure, investment and industry demand. It means backing translation, not assuming it will happen by itself.
And it means involving the workforce early, because the people who understand the work are central to changing it.
It also means being clear-eyed about capability.
Data centres and compute matter.
So do management, trusted institutions, procurement pathways and firms with the confidence to invest. These less glamorous parts of the system are often what determine whether technology produces results.
This is where government choices matter. The task is not to persuade every firm to buy an AI tool. It is to create the conditions in which Australian firms can use AI to solve harder problems, create new markets and compete from here. That means backing research translation, using procurement to create demand, widening access to compute, building skills and sustaining patient investment. These are not supporting details. They are the industrial policy.
Responsible practice belongs in that same conversation. Trust, safety and accountability are not separate from adoption. They are part of what makes adoption possible and sustainable.
The report brings these issues together in a practical way. It does not remove the hard choices ahead. It makes clearer where those choices sit, and why they need to be made together.
At UTS, we see AI as central to Australia’s next phase of industrial transformation, and we intend to help drive it.
Our role is to develop the nation’s skills, undertake frontier research, and connect that research and those skills with the firms, investment and institutions needed to turn it into national capability, not simply greater efficiency, but stronger industries, better jobs and lasting prosperity.
The question is not whether AI will change industry. It will. The question is whether Australia will simply adopt what others create, or use AI to build industries, capabilities and prosperity of our own. Universities can be a critical catalyst, but only if government, industry and research make deliberate choices and build that capability together. That is the partnership UTS is ready to help lead.
This article was first published in The Mandarin.
