Governing in the age of disruption

It is with great pleasure that I present our new white paper, Governing in the Age of Disruption.

This is the first white paper we have published in a while, and a new departure for Symphony3: we have complemented our own graft and insights with AI. Claude has been a new member of the team as researcher, prompter, first drafter and much more.

We published it to coincide with three events in our business in recent months:

All three caused the leaders in our business to think deeply about the implications of AI, both for us and for the organisations and communities we serve. How do we lead in an age where systems are, in the broad sense, more intelligent than we are? How do we keep up when the rate of change is exponential? How do we make sure we have the people, skills and data foundations to take advantage of what is coming, while still building a place where people can flourish? How do we help councils deliver AI that is ethical and human-centred? And in an era where the public sector is squeezed, are there new ways to generate revenue?

This paper asks more questions than it answers. We want CEOs and senior leaders to use it as a tool for thinking deeply about leading efficiently, ethically and energetically in a highly disruptive age. We do not claim to have all the answers, but we do bring experience and diverse skills, and we hope there are perspectives here you can take into your working life.

The body of this document was researched and drafted by Claude, under prompting and guidance from the Symphony3 team, and edited by human beings. Woven through it are articles written by real, experienced people with expertise across technology, business, public sector management and academia. We have tried to combine the intelligence of AI with the wisdom of experienced people, and we hope the result is valuable, insightful and thought provoking.

Kind regards,
Fergal Coleman
Co-Founder and Chief Revenue Officer at Symphony3

Artificial intelligence is rapidly becoming a practical leadership issue for local government, not simply a technology discussion. For LGPro, one of the clearest signals came through our research into the future of the local government workforce: AI literacy is emerging as a critical capability across the sector. At the same time, leaders have lacked a clear, sector-wide picture of how AI is actually being adopted and what is needed to translate its potential into meaningful results.

That is why LGPro launched Australia's first research into AI adoption across Victorian local government earlier this year. It provides an important baseline for the sector, helping us move from assumptions and isolated examples towards informed discussion and more confident decision-making. It gives leaders practical evidence about what is driving adoption, what is holding it back, and how councils can share in the benefits.

In a sector facing persistent financial pressures, growing community expectations, and continuing workforce challenges, productivity gains of this kind cannot be dismissed. However, the research also reinforces that successful adoption is not achieved simply by providing access to new tools.

Different groups of employees are engaging with AI in very different ways. Some are confident and active users, while others are constrained by capability, access, uncertainty, or legitimate concerns about risk. Others are choosing not to participate at all. Each group requires a different response. A single training program or technology rollout will not be enough. Councils need a human-centred approach that builds literacy and confidence, sets clear expectations, maintains appropriate safeguards, and gives employees a meaningful role in redesigning their work.

For CEOs, the starting point is the same as it would be for any other significant investment: a clear understanding of the purpose. AI needs to be connected to organisational priorities, desired outcomes, and measurable public value. This is particularly important when leaders are balancing many competing demands. The strongest case for AI will not be that the technology is new, but that it can help address known barriers, strengthen financial sustainability, improve services, and enable employees to focus on work requiring judgement, empathy, creativity, and connection.

No council should have to navigate this transition alone. Much of the necessary experience and expertise already exists within local government. LGPro's role here is to help connect it. Through our networks, Special Interest Groups, research, and professional development platforms, we are making the knowledge base and collective wisdom that is growing in this space more accessible, reducing duplicated effort, and creating opportunities for councils to learn from one another.

Our ambition is that the sector moves beyond anxiety and uncertainty towards the confident, responsible, and practical use of AI. We want to see the conversation evolve to be less narrowly focused on the spectre of job insecurity and more focused on how AI can genuinely help people perform their roles more effectively and meaningfully. Done well, this transition can be a great enabler, improving productivity, reducing barriers to work, and strengthening local government's position as an employer of choice.

Our ongoing research and partnership with thought leaders across the sector has been a starting point, not a conclusion. Papers like this continue this important knowledge sharing. LGPro intends to build on this evidence base and support papers like this so that they can make the greatest practical difference. By learning together, local government can shape an approach to AI — and the many challenges and opportunities we face as a sector — that strengthens both our workforce and the communities we serve.

AI can now draft your media releases, answer your residents' questions and summarise your board papers in seconds. So what's left for the humans? My answer: the part that was always hardest. The machine can tell you what an answer looks like. It cannot tell you what matters, what's at stake, or who bears the consequences. That is judgement, and in the AI age it becomes your scarcest asset — not your cheapest.

Practical wisdom: what your best people have that the machine doesn't

Think of the difference between a new graduate planner and your best senior officer. On paper they know the same rules. But the senior officer knows which developer will lodge an appeal, which street will pack the gallery at the next council meeting, and when "technically compliant" still isn't the right call. The Greeks called this phronesis — practical wisdom: judgement built from experience, proportion and responsibility. AI has read everything and experienced nothing. It is a brilliant graduate who has never sat through an angry public meeting.

AI smoothness v human friction

AI's great talent is smoothness. Ask it for a response to a community complaint about a pool closure and you'll get something polished, calm and reasonable-sounding in ten seconds. That smoothness is seductive — and it's exactly where the danger sits, because fluency is not evidence. A confident answer about permit requirements can be flatly wrong, and it reads identically to a right one.

The remedy is deliberate human friction. The hour your executive team spends arguing over the wording of a difficult announcement isn't wasted time — it's usually where someone says "hang on, have we actually consulted the residents' group?" The awkward first draft is where thinking happens; skip it and you skip the thinking. And the machine doesn't carry your scars: it doesn't know that the last time council used the word "rationalisation" about a library it made the front page, or why the 2019 depot consultation went so badly. Your veterans know. Keep them in the loop.

Strategic slowness

Because AI makes everything faster, speed stops being a differentiator — and slowness becomes a choice you can make on purpose. I call this strategic slowness. Some things should be fast: rates queries, bin-day questions, routine correspondence. But some situations need hesitation, consultation, or an honest "this is not yet settled" — and AI never produces that answer. It always gives you something clean. A hardship application, a contested rezoning, anything touching a vulnerable resident: these deserve the slow lane, precisely because a fast, plausible answer is available.

You own what it writes

When Air Canada's chatbot gave a customer wrong information about refunds, a tribunal held the airline responsible. The tool cannot absorb liability, and neither can "the AI did it." If your council's chatbot tells a resident they don't need a permit, that is council advice. If a report goes out under your name, it is your report, whoever — or whatever — drafted it. There's a cautionary tale here: a senior university leader published an article defending standards on AI use, and it was later removed when it was revealed that AI had assisted in its drafting. The reputational damage came not from using the tool but from the gap between what was claimed and what was done. Use AI openly, disclose it where it matters, and never let disclosure substitute for ownership.

The optimistic case

None of this is anti-AI. From the printing press to the internet, every wave of technology has democratised knowledge, triggered a panic about human capacity, and ultimately given us new capabilities. AI, well overseen, can be a patient one-on-one explainer for every member of staff and every resident — that is a profound opportunity for local government. The machines will keep getting faster and smoother. That is precisely why the value of slow, accountable, experienced human judgement is rising, not falling. The job of the CEO is to make sure the council still has it.

The need for change

When I became CIO of the Kmart Group, I brought together, for the first time in Kmart or Target's history, the IT, advanced analytics, engineering, data and product management teams. Together we supported 50,000 employees in more than 500 stores, delivering more than ten million customer interactions every week and selling one billion units each year. One of the first questions I was asked was: "Who is our biggest competitor, and what will we do about it?" My answer was not what anyone expected. Our biggest threat was our own customers' expectations.

Customers do not compare your organisation only with organisations that look like yours. They compare every experience with the best experience they have had anywhere. Netflix changed expectations around personalisation. Uber changed expectations around convenience and visibility. Amazon changed expectations around speed. Those expectations do not stay inside those industries — they are boundaryless, and they follow customers everywhere.

A council might not consider DoorDash, Netflix or Airbnb to be in the same wheelhouse, but they are certainly influencing how your citizens view you and the services you provide.

Your benchmark for the next great experience is your last great experience. Once you start paying with your phone, cash feels inconvenient, and the council's multi-step web form now feels unacceptable. When you can get shopping delivered within hours, waiting a week for a council response feels archaic. The benchmark shifts instantly, and constantly. Seeing a car approach on a map changes expectations for all transport services — and for workflows within councils. Uber users want to know: where is my planning application up to? Netflix users want council websites to be simple, immediate, smooth and personalised. Recommendations for what's next would be nice too.

If your council service feels slower, harder or more confusing than your citizen's last great experience, you lose. To win, councils must embrace AI and design for the rising bar set by others — because your customer brings their best experience with them every time.

You cannot transform an organisation without transforming its people

Technology changes faster than you can hire. ChatGPT was released to the public just before Christmas 2022. Claude Code became generally available in May 2025. You cannot hire someone with ten years of Claude experience, or someone who used it through a four-year degree — they do not exist. So you either continually hire the latest and greatest, or you help your team learn the skills required.

At Kmart I chose the latter. The skills we have today do not match the jobs of tomorrow. When I became CIO, I decided to train every single Kmart and Target IT person in cloud computing and get them certified in AWS. This was an ambitious undertaking: not a single company in the world had done it at this scale, not even Amazon itself. The most exciting outcome, apart from being first in the world, was the change in mindset we saw in the team. When a grandmother who last sat an exam in 1979 obtains a cloud certification in 2020 — and talks to her grandson about Fortnite running in the cloud — you have made a significant change to how staff see the business and solve problems, and opened doors for that individual that will never be closed. We had not simply taught someone a new technology. We had changed what that person believed they were capable of doing.

When I became Chief Transformation Officer at Village Roadshow, I took that concept one step further. With AWS, we created a 12-week gamified cloud-learning competition called the Guild Tournament. Teams learned together, collaborated across functions and submitted ideas to solve real business problems. The winning idea was actually built, and a lot of fun was had along the way. Everyone learned and upskilled, and the organisation got a batch of new ideas, stronger collaboration and greater confidence in what its people could achieve. The objective of transformation should not simply be to implement new technology. It should be to leave the organisation more capable than you found it.

Don't go it alone

"What is your biggest failure?" It is the interview question most people dread, but learning things the hard way cements the lesson. My biggest failure as Kmart's CIO is that I did not bring in outside help soon enough to accelerate our cloud journey. My intentions were noble: I wanted to control costs, build our skills internally and avoid solving a capability problem by simply hiring more people. But the gap in skills and capability was too big, we moved too slowly, and I did not bring anyone in from outside for a long time. Eventually I realised the error of my ways and hired Digi-Ren, AWS specialists. The effect was almost immediate — they provided the spark and the backstop that supercharged our cloud efforts, coached our team, and left us with skills that were self-sustaining.

That experience changed my view of leadership. You do not have to solve every problem yourself. The important thing is knowing where your capability ends, recognising the gap early, and bringing in the right expertise before the gap becomes a constraint. Good external support should not create dependency.

Help is always within reach. The key is knowing when you need it, who can provide it, and having the judgement to ask. No leader, team or organisation has every capability it needs at every point in its journey. The strongest leaders recognise the gaps early, bring in the right expertise and use it to build capability that remains long after the external support has gone. There are people and organisations ready, willing and able to assist — you already know some of them. You just have to ask.

Michael Fagan is Strategic Advisor to Symphony3. He was previously Group Chief Information Officer of the Kmart Group and Chief Transformation Officer of Village Roadshow.

"What is the city but the people?"

Shakespeare was right. People are, always have been, and always will be the backbone of Australia's communities, culture and quality of life.

To each other, in our communities, we provide care, love, friendship, support in tough times, essential services and a better quality of living. And over time, technology has enabled us to do this in ways we never would have imagined earlier in our lifetimes. Whether it be computerised library catalogues, online rates payments or smart rubbish bins, advances in the tools provided to our council workers have allowed them to give more back to our people.

And so, just as I tell my clients in business and federal government, councils must think of AI as the next step in empowering people to do more and give more back to our communities.

AI must be seen as a workforce opportunity, not a challenge.

This is timely, because the financial and workforce pressures on our councils are arguably the hardest they have been for decades. Budgets, and the salaries that can be afforded for the council workforce, are limited. In Victoria, 62 per cent of responding councils said that salaries and remuneration were the most significant driver of occupational shortages. And to add to this pressure, many skills essential for councils to run effectively are already in significant shortage: 81 per cent of responding Victorian councils reported occupational shortages in 2024–2025, in roles such as building surveyors, engineers, and urban and town planners.

But my time working with corporate and federal government clients has shown that this does not need to be a constraint.

AI will change the skills required, the ways of operating, and the breadth and depth of the work your teams can do. This will enrich the lives of council workers, not take away from the quality of their work.

And so the biggest constraint councils should worry about is not financial or occupational. It is emotional: fear. Or, more specifically, the fear of embracing AI when we need it most.

Panic makes people unproductive, more cautious and more defensive. Academics from Wharton and Boston have argued that many avoid AI because it threatens their expertise and autonomy. Others will simply choose to switch off. For Australia this is particularly pronounced: the University of Melbourne last year found Australians ranked the least optimistic towards AI of 47 surveyed countries.

From retirees to graduates, to mums returning from parental leave, to software engineers, to the most senior and successful executives in business, technology and strategy, almost no person I have ever spoken to is immune to worrying that their job will be taken.

Given the daily barrage of newspaper headlines carrying bold and unproven claims from tech giants that mass job loss is imminent and existential, we cannot be blamed for worrying. But that worry turns into a second job of its own. It tends to make people focus on self-preservation, resist change and worry about internal politics.

We must instead channel Roosevelt and remind ourselves: "the only thing we have to fear is fear itself".

AI provides councils, as it does all my clients, an opportunity to innovate and reimagine how and what is done for our country and communities.

But let's be clear on the facts of why, so you are armed with them for your conversations with your employees.

Firstly, a job is not a task. A job is a responsibility.

A maternal and child health nurse does so much more than weigh a baby. They sense how a mother is feeling, and judge how best to discuss a lack of weight gain in the baby. They take the weight history and combine it with context and inference to judge whether the baby needs medical help. The more that AI takes notes, records data and provides checklists, the more the nurse can focus on checking her intuition and considering how best to engage her patients on challenging topics that, in that moment, she — not the technology — is responsible for.

Secondly, technology on its own does not create change.

A London School of Economics study of police departments in the United States between 1987 and 2003 found that when departments used computerised record keeping they saw reported crime rise by about 10 per cent, because it made crimes easier to report. However, when that technology was combined with a broader management system that used crime data to deploy officers, solve problems and hold managers accountable, the proportion of crimes resulting in an arrest rose by approximately two percentage points — equivalent to roughly a 10 per cent improvement.

AI gives organisations — and councils are no exception — the opportunity to reimagine their operating model.

Counter to what we are frequently told, it is not hard to use either. Nor are you and your people too late.

In fact, the fundamental skills that make us good in the workplace are the same skills that make us good with AI. When we are curious, we experiment with AI. When we have good judgement and high standards, we know when to use AI to speed up a process and when we should not. When we bring IQ, we know when to question and how to probe the findings it gives us. When we have EQ, we know what will resonate and what will not. And most importantly, when we are confident, we know where and when we do things better than AI.

And so, with confidence, your people can quickly learn to adapt to AI, and to excel at using it to deliver better outcomes for your community.

Teaching those skills also makes councils a better place to work.

When a Stanford and MIT study looked at more than 5,000 customer service operators given an AI assistant, productivity rose by an average of 15 per cent. By putting the knowledge of the best performers within everyone's reach, those with the least experience gained the most.

When it is hard to pay high salaries — and even when it is not — I tell my clients to give their employees purpose, direction and growth. For a council, purpose exists in your very nature. To provide direction, you must quell any fear of job loss from AI and map out the path ahead.

And to grow your people, and to continue to evolve into the community and the city that Shakespeare spoke of, give them the tools, show them how to use them, and, importantly, tell them not to fear.

Together with your people, AI will help your cities grow.

Dr Vivienne Groves is Managing Director of Ero Co, a management consulting company advising business and federal government on strategy, operating models and AI adoption.

On average, around 40% of Australian local government revenue comes from sources other than general rates and utility charges; excluding grants, user charges and fees typically account for around 25% of council income. This is a larger — and more controllable — revenue base than most executives appreciate, and unlike capped rate revenue it is one a council can grow through its own effort. Yet it is actively, commercially managed in only a small minority of councils.

Revenue management is the discipline of deciding what to offer, to whom, when, and at what price — using data and segmentation rather than a static annual fee schedule. It transformed airlines, hotels and media; councils, already running aquatic centres, venues, carparks, childcare and waste services, are a natural if late addition. The levers are pricing, inventory, marketing and channels; the process is data collection, segmentation, forecasting, optimisation and continual re-evaluation.

Australian councils need to promote their services more consistently so residents know what is available and how to access it. Councils already own trusted local channels — websites, newsletters, facilities and community networks — but often use them unevenly. A more disciplined approach to community messaging would improve service awareness and could also attract appropriate government campaign funding or ethical sponsorship. Any paid content should meet clear public-interest standards, and councils should retain the right to decline or withdraw it. Revenue raised this way can reduce pressure on rates, especially for pensioners and fixed-income households.

The barriers are real — cost-recovery legislation, competitive neutrality, political will, and the 'front page' risk of a poorly communicated price rise — which is why staging matters. The starting point is the call to action above: know every detail of every line in your revenue portfolio, and commission a stage-one revenue and cost-basis review ahead of the next budget cycle.

This article is condensed from a companion paper prepared by Anthony Dow, author of Local Government Revenue Generation.

This paper asks council CEOs some uncomfortable questions about the decade ahead. It is only fair that we answer a few of them ourselves. Symphony3 is a small, Australian-owned technology company. We are not immune to the forces described in these pages — we are living them. So rather than offer advice from the sidelines, I want to share what leading through this actually looks like from inside a technology company, because I believe the lessons transfer directly to leading a council.

Same purpose. Same customers. Same tagline. A different way of working.

This paper asks what will change in the next ten years and what will not. We put the same two questions to ourselves, and the answers now shape how we run the company.

What will not change for Symphony3: our vision, our customers, our purpose, and our tagline — simple, connected customer experiences. Councils needed that before AI arrived and they will need it long after the current wave of tools has been forgotten. Those constants are settled, and we protect them.

What must change is how we deliver them. Technology now demands that we do things differently — and this is not optional. A technology company that keeps doing things the way it always has will eventually be extinct. It will also fail in a quieter duty: the duty we owe our people to help them develop and modify their roles for what is coming, rather than leaving them stranded in roles the market no longer values. We take both seriously. Survival and stewardship point in the same direction.

We work on the new business model every day

The most important shift we made was treating this as daily work, not a transformation program with an end date. We are working every day on what the new business model is — testing it, adjusting it, and building it into our business plans with the same discipline we would apply to a major product investment. AI done ad hoc produces interest but not results. We know, because we tried some of that first.

The practical engine of this is our intern program. We have brought five young interns into the business and partnered one with each of our product lines and with our business support function. Their job, alongside our experienced people, is to hunt for quick wins and automation — and to help us build new checkpoints into how we work. Fresh thinking from people born after the smartphone, paired with the deep domain knowledge of our existing team, deliberately breaking down the old barriers between functions. Neither works without the other.

The investments are specific

None of this is abstract. The test we apply inside the business is simple: is AI reducing what it takes to build what we build? Our current investments are deliberate and concrete: tools and research that let us build and migrate websites more quickly; training our customers with content developed during the build phase, rather than writing it after the fact; fast-tracking how we build connectors between council systems; and automating testing. Each one drives down the cost and time of delivery, and each one was chosen because the benefit is measurable, not because the technology is fashionable.

The principle this paper puts to councils — redesign the work first, then apply the technology — is the same one we apply to ourselves. AI amplifies good processes and exposes bad ones. We have found no exceptions.

The question to ask every vendor — including us

AI is also now part of what we sell — our Beetrix AI agent reviewing parking infringements at Warrnambool City Council is one example of AI moved from pilot to production. But we do not add AI to our solutions because it is fashionable. We add it where it strengthens the things that make us genuinely different, because the AI age will punish the undifferentiated. When anyone can generate software quickly, the value shifts to what cannot be generated — trust, domain knowledge, integration depth, and relationships built over years of service.

We are happy to be asked, because the principle is not new for us. Our SmartGlue integration platform was built around reuse: every connector we build for one council becomes available to the next, so each deployment gets faster and cheaper than the one before it. AI is now accelerating that same principle — and for us, honestly, that part is a work in progress, measured as we go rather than declared as done. Treat any vendor who tells you it is solved with more suspicion, not less.

A closing word to council CEOs

Start where we started: name what will not change. Your community, your democratic mandate, your purpose — these are your constants, and they are stronger anchors than ours. Then be honest that everything about how you deliver them must be open to change, and that this is now daily work for the whole organisation, not a project on the side. I know your constraints are heavier than ours — enterprise agreements, elected members, and a community watching every change. The direction is the same; only the pace differs.

And treat your people as the point of the exercise, not the obstacle to it. Helping people develop and modify their roles for the AI age is not the price of transformation — it is a duty of leadership, and the organisations that honour it will keep the trust and the talent that the others lose.

John Nevins, Local Government Advisor, Strategy & Digital, describes the leadership task well: management gets things done, while leadership creates the environment in which people and the organisation can be at their best. For councils adopting AI, that means giving employees the tools, training and space to learn how to apply it in service of organisational purpose.

The opportunity in front of councils is the same one in front of us. It will not be captured by waiting, and it will not be captured ad hoc. Same purpose, different way of working — worked on every day. That is what we are doing, and we are happy to compare notes along the way.

Thomas Hynes is Chief Executive Officer of Symphony3, an Australian-owned technology and integration company serving local government across Australia and New Zealand.