Accelerating AI innovation in government

The Business Council of Australia recently warned that Australia risks falling behind other nations in AI innovation without a clear, government-led national AI strategy in place. It called for urgent action to fast-track skills, infrastructure and innovation to position the country as a global leader by 2028.

This puts a spotlight on government leaders, who are already under pressure to accelerate their AI initiatives regardless of tight budgets. While many governments have laid the groundwork with initial AI policies and adoption plans, few have formalised a comprehensive strategy at the departmental or agency level.

According to a Gartner survey of 499 government CIOs and technology executives, the top barriers for scaling AI adoption include persistent concerns around AI ethics, fairness and bias, along with challenges in building trust in AI models and estimating the value and benefits. As a result, progress is slow and efforts to move beyond pilots are often poorly structured. But it’s clear that AI adoption will be governance-led.

The challenge for governments is keeping pace with the rate of change. The advent of AI agents is promising a new era of capability. But these systems, which are capable of autonomous decision-making, planning and execution, pose new complexities for the public sector. In the wake of the Royal Commission into the Robodebt scheme, fears around automation are high, potentially stalling the adoption of what could be a truly transformative technology.

Every government department must develop its own unique AI strategy regardless of its ambition or plans for adoption. While more conservative departments may feel unprepared, a well-crafted AI strategy sets realistic expectations for executives and provides a clear roadmap that focuses limited resources on the highest-value opportunities.

The strategy must align with whole-of-government priorities, drive cross-department collaboration and link AI investment to leadership priorities around mission, efficiency and effectiveness. It must also deliver measurable business value.

Define the AI vision

A well-defined AI strategy is about setting the vision and goals of what a government agency or department wants to achieve with AI, in alignment with the overall organisational strategy.

An AI vision should define value and support priorities at a departmental level, such as improving citizen experience, increasing workforce productivity and managing risk.

Establish governance for AI adoption

With a clear AI vision, governments can make more deliberate and strategic technology investment decisions. But success will be directly linked to public trust and acceptance of how AI is adopted and used. There is growing demand for safe, transparent and responsible AI. According to a recent Gartner survey, 89% of government CIOs and IT leaders cited responsible AI as a top priority.

An AI strategy will outline the governance frameworks that will build trust, manage risk and ensure long-term success. The first generation frameworks have been published at both a federal and state level, but need to be incorporated into the way of working for every department and agency.

Even the most conservative government organisation needs a framework that ensures existing and new vendors and partners are aligned to its AI vision and responsible AI guidelines.

Align on AI ambition

Governments must clearly define their AI ambition. This will set the pace and direction for how fast and far they’re willing to go with AI adoption. This isn’t a technology challenge: it requires close collaboration with senior leadership to ensure initiatives are aligned with broader organisational priorities.

Today, many governments frame their AI ambitions around productivity gains. However, much of the early activities are focused on internal-facing initiatives, such as improving back-office operations, rather than delivering direct benefits to citizens.

Generative AI, in particular, is proving to be a common starting point, with 56% of government CIOs in Australia and New Zealand reporting their department or agency has deployed or is in the process of deploying initiatives in 2025.

Link goals to AI priorities

The next step is to set priorities that align closely with the agency’s mission and objectives. This approach helps governments build a structured portfolio of AI use cases. By mapping these against business priorities and expected value, government CIOs can better assess where AI can deliver the greatest impact and allocate resources accordingly.

For example, if the focus is on excelling in citizen experience and strengthening trust in government services, AI initiatives should prioritise improving how citizens access and interact with information. This might involve deploying conversational agents to make it easier for citizens to access accurate information or use summarisation capabilities to deliver more personalised support.

Communicate AI progress to key stakeholders

Proactively communicating the adoption and use of AI is critical to building trust with key stakeholders. It’s important to highlight the implementation of AI assurance frameworks and governance structures, particularly those designed to safeguard sensitive information, such as citizen data and privacy. Demonstrating a commitment to responsible AI use helps build confidence across both internal and external audiences.

Governments should provide clear updates on the benefits being realised, while also outlining the measures in place to test and monitor to ensure AI stays within the established guiderails.

Equally important is investment in employee awareness and capability. An AI strategy should highlight the organisation’s journey as AI becomes more embedded in government operations. Governments must foster a culture of understanding and readiness among employees to ensure investments in AI are successfully incorporated into the standard ways of working across the workforce.

*Dean Lacheca is a VP analyst at Gartner, who will be presenting on how governments can successfully deploy and sustain AI at the Gartner Data & Analytics Summit, 17-18 June in Sydney.

Image credit: iStock.com/MF3d