Agentic AI could be Australia's productivity superpower if we seize it
Reversing Australia’s weak productivity performance is one of the most urgent economic challenges facing the re-elected Albanese government.
On the current trajectory, Australians could be 40% poorer and will work 5% longer weeks by 2063. Left unaddressed, declining productivity risks not only our nation’s economic strength but also the quality and efficiency of the services that the government provides to all Australians.
While it is positive to see Treasurer Jim Chalmers make it his second-term mandate to lift productivity by kicking off a new round of consultations for reform — which includes ‘enabling AI’s productivity potential’ as a priority — the Australian public sector faces inherent challenges that threaten its progress and ability to lead by example.
Legacy systems, outdated infrastructure and siloed processes remain widespread across government. Coupled with a persistent skills shortage, it’s impacting how quickly the public sector can modernise and adapt, which is key to being able to deliver better and higher-quality services to the citizens of Australia
Fortunately, the arrival of agentic AI — the next phase of the AI revolution — presents an opportunity to get our nation back on track. AI agents are already transforming our approach to work in the private sector, and it’s a technology we can harness right now to drive tangible productivity outcomes in the public sector.
What are AI agents?
Unlike traditional automation, AI agents go beyond task execution; they can orchestrate and manage entire workflows. In other words, they don’t wait for instructions — they proactively plan and execute tasks independently, adapt to feedback, and use external tools and data, allowing teams to move faster and multiply their output when it comes to rudimentary tasks.
According to Microsoft’s Work Trend Index 2025, 81% of business leaders globally expect AI agents to be part of their strategy in the next 12–18 months. Industry leaders like EY, Carvana and the Commonwealth Bank of Australia are already capitalising on the impact of agentic AI, achieving productivity leaps and positioning themselves at the forefront of their sectors. Imagine what this could mean for government agencies: faster delivery of digital services, more cost-effective project management, and a public sector workforce empowered to focus on higher-value problem-solving and citizen-centric work.
Redefining productivity for government
Large government IT projects are notorious for cost overruns and delays and, as highlighted by the Productivity Commission, successfully scaling innovations across diverse public sector needs is challenging due to scarce resources and varied local contexts. Agentic AI offers a powerful lever to address this challenge as the technology moves from a prompt-based tool to a deployable resource.
By automating routine processes and reducing manual errors, government agencies could accelerate project timelines, optimise resource allocation, and ultimately deliver better outcomes for Australians. In software development, for example, agentic AI tools effectively act like an extra teammate — able to draft pull requests, run tests, fix bugs and refactor code. This saves developers hours each day, shortens release cycles and helps teams ship faster.
The time and money saved could then be reinvested in innovation and public value, rather than patching up costly overruns. This renewed focus on innovation presents a significant opportunity for the public sector to attract top IT and developer talent by embracing the next wave of agentic AI capabilities — and to retain key personnel with critical systems knowledge.
A recognised shortage of specialist skills in domains like ICT and cybersecurity within the public service is also impeding technological adoption and progress. While the government cannot control the pipeline of tech graduates passionate about public service, there is an opportunity to leverage agentic AI to amplify existing talent, boosting capability from within.
Security, trust and governance
Of course, innovation must not come at the expense of public trust or security. People are wary of how their data is being handled and used; scepticism could increase with AI use in the public sector. This makes it vital to be transparent and to have strong governance and frameworks in place.
However, public sector agencies needn’t compromise on security to benefit from AI agents, but they should hold the tools they use to a high standard. For example, tools like coding agents should be designed specifically to ensure an organisation’s existing security posture remains rock solid: that branch protections, repository rules and organisational policies continue to function as expected.
Every pull request created by an agent should be met with human approval before any workflows run — and this should be consistent for any tool, not just those designed for software development. Further, modernisation through agentic AI also ensures that legacy systems are more secure and futureproof, accelerating the transition to safer, more resilient infrastructure.
Australia cannot afford to squander AI’s potential
Australia has the talent, the ambition and the digital foundation to be a global leader in agentic AI. If we hesitate, we risk falling behind in the global race for digital leadership. The consequences? Slower service delivery, rising costs, and missed opportunities to build a more engaged, innovative public sector workforce — all while other nations unlock new levels of efficiency and public value.
Australia’s public sector can and should lead the way. By championing AI adoption and upskilling our workforce, government can set the standard for a new wave of meaningful digital transformation, delivering more responsive, resilient and cost-effective services. Agentic AI could be a turning point to reverse a decade-long productivity decline for citizens today and future generations, but only if we seize the opportunity.
Policy to practice: simulations can bridge the government AI skills gap
While the private sector seems to be embracing AI 'smarts', many government employees...
Interview: Phil Zammit, Avaya
For our summer Leaders in Technology series we are discussing AI, system complexity, skills and...
From training to collaboration: how XR can transform public sector operations
Extended reality (XR) has emerged in recent years as a powerful tool for public service...
