Australia’s next AI frontier is on the job site

By Steven Miller, Area Vice President, Microsoft ANZ

The following piece was originally published as an editorial in The Australian on 27 May, 2026.

Hands-on roles are often seen as more resilient to technological change than office-based knowledge work. Yet there’s been little conversation on how AI could improve the daily work of field workers, leaving a missed opportunity to boost productivity, as well as job fulfillment.

Australia is facing a persistent productivity slowdown. A Productivity Commission report found Australia’s labour productivity growth was at its lowest in six decades, well below the OECD average, with field-based industries like construction lagging the most. AI is widely discussed as a lever for productivity, but an untapped opportunity lies in enabling the nearly two million skilled tradespeople who keep the country running.

This is where the next wave of AI is, and where some of Australia’s toughest productivity challenges persist. For example, construction industry labour productivity grew only about 17% in the past 30 years, compared to a 64% market-sector average. Workers in this sector want to be “on the tools,” applying their craft, rather than buried in paperwork and admin. This is where AI can make a meaningful difference. Used effectively, it can shift what field workers spend time on, improve safety outcomes and ultimately drive economic growth.

Yet few Australian companies have a plan to seize this field-AI opportunity, and no single group can manage it alone. Too often, AI is introduced without the early, meaningful consultation that builds trust or translates into sustained adoption. Since January, Microsoft has been collaborating with the Australian Council of Trade Unions to elevate workers’ voices in AI adoption. Many of their affiliated unions represent tradespeople in the field and on job sites.

Microsoft’s work with the ACTU reflects a growing recognition that success with AI will not be achieved with technology capability alone, but on whether it delivers fair and inclusive outcomes for workers and society. This work supports the National AI Plan, aligning AI investment with public trust and responsible deployment.

For site-based teams, AI has the potential to act as a single interface by handling routine processes in the background. Imagine a supervisor saying to an AI agent, “I need to report an incident on site A,” and then, still with human oversight, AI completes the required forms, notifications and system updates. By reducing the administrative burden, AI could also help address one of the biggest staffing constraints these industries face – a shortage of skilled supervisors to oversee job sites.

However, simply adding another system or app is not the answer. Too often, technology is pushed onto field workers without proper consultation, while also creating a patchwork of systems that do not talk to each other.

Some Australian organisations are showing what is possible. We have customers who have taken an inclusive approach from the start. When I was visiting an energy company recently, I spoke to an experienced technician who shared that he used to spend hours after work writing reports, and touch typing wasn’t something he ever learned. Now he simply dictates his notes and uses Copilot to transcribe and polish them. After a quick review, he sends off his reports in minutes, freeing up hours to focus on his craft.

We have also learnt that adoption in the field requires a different approach to skills and change management. It needs tailored training, supportive leadership and technology that works securely in remote environments. Most importantly, new systems must respect how people work. Trades have deep expertise and proud traditions. AI should support that knowledge, not override it.

Training provider Akkodis Academy has partnered with Microsoft and customers like Rio Tinto, to tailor learning modules around how field and trade-focused employees would use AI in practice. These modules account for job-specific language, acquired shorthand, and a greater reliance on voice-based tools over typing.

Infrastructure, resources and services leader CIMIC Group is another great example we’re working with. They are applying AI across major engineering projects in Australia and globally. Using AI-enabled solutions such as digital twins, predictive analytics and Copilot agents, teams are gaining earlier insights while reducing reporting, admin, and supporting delivery certainty. That means more time on work that enables them to better operate safely, manage risk, improve sustainability outcomes and lift productivity on site.

To scale these approaches, three priorities stand out. First, invest in practical, industry-led AI training that is purpose built for field workers. Second, give workers a genuine voice in how AI is rolled out and adopted. Inclusive design is imperative for social acceptance and trust in AI. Third, celebrate real success stories, because nothing sparks confidence like seeing how a mate made it work.

Australia’s next productivity leap will not come from offices alone. It will come from empowering workers on every site and in every trade with technology that helps them do their best work. If we get this right, AI can make jobs safer, more rewarding and more productive, and ensure no one is left behind.