A public sector challenge: the shift from digital to intelligent services

For more than a decade, Australian government agencies have invested heavily in digitising public services across key areas such as licence renewals, electronic applications and access to information. While this has delivered enormous benefits for government and consumers alike, it does not necessarily transform what happens to the processes behind it.

Even in the current AI-led economy, a citizen may submit a form online in minutes, only for the request to enter a complex chain of manual reviews, emails, data entry, approvals and hand-offs across different teams and systems. At the same time, public sector employees often still need to move information between legacy applications, which can create data risks.

The challenge is no longer simply how to develop more digital services. It is how to turn them into intelligent services capable of understanding context, coordinating work across systems and taking appropriate action, while keeping people in control of the decisions that require human judgement.

From digital government to intelligent services

Many government agencies still have decades of technology investment behind them, with critical information spread across modern cloud applications, bespoke platforms and legacy systems. They must also operate within strict requirements around privacy, security, accountability and administrative processes.

Replacing all those systems is neither practical nor necessary. Instead, the immediate opportunity is to connect them more effectively. This is where business orchestration demonstrates a clear solution; by coordinating AI agents, robots, people and existing technology so they work together as part of an end-to-end government process.

Through business orchestration, an AI agent could interpret an applicant’s request and supporting documents. Automation could retrieve information from authorised government systems and perform routine checks. Another agent might identify missing information or anomalies. Straightforward administrative steps could progress automatically, while complex or sensitive cases are routed to an experienced public servant. The employee is not removed from the process. Their expertise is used where it has the greatest value.

Business orchestration is the layer that coordinates this work: determining what needs to happen next, which system or agent should perform it, when a human needs to become involved and how the entire process is monitored and audited.

Without orchestration, government agencies risk replacing today’s fragmented application environment with tomorrow’s fragmented AI environment. This could result in literally hundreds of AI agents operating across multiple departments without sufficient coordination or visibility of what they are accessing and the actions they take.

Australia is ready to orchestrate

Australian public sector organisations are already demonstrating what becomes possible when repetitive administrative work is automated and employees can concentrate on higher-value services.

At Gold Coast Health in Queensland, for example, automation is returning more than 40,000 hours annually to frontline patient care. This demonstrates an important principle for government transformation: the greatest value of automation is not necessarily the transaction it completes — it is the capacity it returns to the organisation.

Traditional automation excels when rules and processes are clearly defined. AI agents can add the ability to interpret unstructured information, reason about a task and adapt their actions towards a goal. Combine those capabilities into an orchestrated workflow and government can redesign entire services for greater efficiency and an enhanced citizen experience, rather than simply accelerating individual tasks.

Intelligent government must be accountable

For the public sector, speed and productivity cannot be the only measures of success.

Government decisions can affect someone’s income, licence, health care, business or access to essential services. Citizens need to know that decisions are made fairly and that government can explain how those decisions were reached. This makes governance an architectural requirement, not something that can be added after AI is deployed. Agencies need clear rules governing what an AI agent can access, what actions it can take autonomously, what must be escalated to a person and how those actions are logged and reviewed.

Instead of allowing autonomous agents to operate independently across an agency, orchestration provides a mechanism through which organisations can establish boundaries, manage exceptions and keep people involved at critical decision points. This requires agencies to look beyond individual tasks and examine complete processes: where work begins, which systems it passes through, where bottlenecks occur, when decisions are required and where human expertise genuinely adds value.

Australia has already built much of the digital infrastructure needed to make government easier to access. The next challenge is more ambitious: making government easier to operate. If the first era of digital government was about putting services online, the agentic era should be about connecting the work behind those services so agencies can respond faster, use resources more effectively and deliver better outcomes for citizens. Business orchestration makes this possible. Top image credit: iStock.com/davidf