Used well, AI can be a powerful tool that water organisations can harness to deliver on the intent of the water reforms, which is to provide New Zealand communities with water services that are safe, reliable, and affordable. Jarred Griffiths, our Lead for AI Transformation, examines the opportunities and how to make the most of them.
It would be an understatement to say change in the water sector has been relentless over the past five to six years. This July, five water service companies were established, and next year another 13 will follow. Organisations are being established quickly, under a lot of public scrutiny, and with increasing interest from regulators.
The regulatory framework that’s taking shape is designed to lift the standard of planning, asset management, and performance, and to provide more accountability right across the sector. The question worth asking early on is how the tools now available to water organisations, including AI, can help them achieve those goals.
The immediate focus on operational readiness is necessary
In this context, it makes sense that new water organisations are focussed on being operationally ready. They need to make sure that, from Day 1, the water runs and maintenance programmes continue without pause.
We know this is complex work and it needs close attention. These new companies are inheriting assets, data, contracts, and often people from multiple councils, and integrating all this requires a structured approach through until launch day and well after.
.jpg)
But reform also opens a window we can’t afford to miss
However, as well as that short-term priority, the new companies need to keep sight of the opportunity the water reforms create for a more fundamental, longer-term transformation.
Identifying that transformation opportunity early on ensures it doesn’t fall completely off the short-term agenda, and sets it up to be a North Star to orientate strategy development and early investment choices.
Early decisions about systems, data, and processes can set the new water organisations up for the future service model they want. If the longer-term opportunity is considered early, then the initial activity to set up the new organisations can enable future work, rather than baking in systems and practices so that it’s much harder to change course further down the track.
AI is one of the tools that can help achieve the purpose of the reforms
One of the opportunities new organisations should be considering early on is the significant potential for AI to transform how water services are delivered.
The pace and scale of AI development means this is no longer an abstract future opportunity. Generative AI moved from novelty to mainstream use remarkably quickly, and the level of investment now being poured into AI infrastructure is enormous. The technology is improving quickly, spreading across more tools and business processes, and already changing the way work is done across most sectors.
With water services, many of the key business processes are highly amenable to AI. Managing complex assets, leveraging large volumes of data, automating repeatable operational processes, speeding up planning decisions, supporting field work, and supporting regulatory reporting – they’re all areas AI can support.
AI can help the new water companies to make better use of information, to strengthen their operational performance, and to build more resilient and responsive services over time. The fundamentals of service delivery will remain, but the way work is done could potentially be dramatically different. Now, as the new entities are being formed, it’s the right time to look hard at current processes and ways of working and to start to reimagine them.
International examples show what better outcomes can look like
Some compelling overseas examples are giving a sense of a possible AI future for water and other key services, especially the scope to improve how assets are managed.
The Government’s view is that the wider public sector has a poor track record of asset management, with numerous high-profile failures and a growing evidence base of unacceptable asset-management practices. In the context of that pressure, the overseas examples can help our water sector respond and do better.
In drinking water networks, AI is being used overseas to detect leaks and identify unauthorised use. In Queensland, Unitywater’s deployment of AI has reportedly detected more than 10,413 ML of leakage and unauthorised use since 2013, equivalent to AUD$27.9 million in savings. In London, a Thames Water project worked with technology vendors to monitor 350 km of network and saved 2,376 ML of water in its first year.
The same pattern is evident in wastewater and treatment plant operations. In South Bend, Indiana, AI and smart technology have transformed the management of overflows. Smart sewer technology, sensors, and real-time controls have helped to reduce combined sewer overflow volumes by more than 70%, avoiding around one billion gallons of overflow each year. This has helped defer the need for hundreds of millions of dollars in traditional capital works.
In a number of treatment plants around the world, the technical services and engineering firm Jacobs has been successfully using AI and machine learning to recommend chemical dosing, aeration, and operating setpoints. In case studies, plants adopting this approach report chemical savings of 10% to 30% while continuing to be compliant.
These concrete examples show how service delivery can be transformed through better use of data, providing an opportunity to reduce capital spend and support staff to make better decisions and improve operations – all outcomes that our water reforms are trying to achieve.
There are wider opportunities to explore and scale up AI benefits at home
AI can strengthen asset management by turning fragmented data on assets, maintenance, telemetry, GIS, and operations into better intelligence for planning and investment.
Pipe failure models can help prioritise inspection and renewal programmes. Pump analytics can identify early signs of failure before emergency call-outs are needed. Digital twins can bring live and static data together to support operational decisions, emergency response, and lifecycle planning.
The common thread is that AI tools can improve the management and long-term stewardship of assets. They can give better visibility of risk and stronger evidence for investment choices. They can also allow more targeted maintenance, helping operators to shift from reactive fixes to more predictive management of critical infrastructure.
Beyond infrastructure and operations, AI can potentially help with customer service and community engagement. From intelligent virtual assistants and proactive outage notifications, to demand forecasting and personalised communications, AI can help organisations respond more quickly, improve service quality, and build stronger customer relationships. Those applications may seem less tangible and central than innovations in managing assets, but they can have a direct impact on customer experiences and public trust.
The low quality of data will make it harder to realise large-scale benefits
For many leaders, the promise of AI is now easy to see. The harder task is translating that into large-scale use, sustained capability, and measurable benefits. New water organisations should approach this as an operating model challenge, not simply a technology choice.
We should acknowledge up front that complex high-value applications of AI like the examples I’ve described are unlikely to be achievable in the short term. This is partly because new water organisations are inheriting low-quality data in many respects, particularly around the condition of assets.
Data maturity tends to be low right across the wider public sector. While organisations hold vast amounts of data, it’s often not managed as a strategic asset – that is, it’s not structured in a way that allows it to be fully utilised, and it’s not integrated across core business processes.
AI can only be as useful as the data and processes it draws on. For example, leak detection, predictive maintenance, digital twins, and renewal planning all rely on accurate asset registers, integrated telemetry, consistent maintenance histories, and clear operational processes and workflows.
If we advance AI use cases without a strong data foundation, this risks automating current gaps, inconsistencies, and workarounds. In that situation, the technology is unlikely to prove its value.
That makes data and technology core enabling infrastructure
This is where it will be important for newly formed water organisations to consider how to get ahead of these risks, and position themselves for a longer-term transformation. The priority must absolutely be data management and data capability.
In practical terms, this means the new organisations should be deliberate about the data capability they build from the start. They need a clear data strategy that identifies the data that matters most, how it will be governed, and how it will support decision making around operations, assets, and investments.
The new water companies also need the capability to make use of that data. So as well as investing in modern platforms, this means investing in people who have the right technical and analytical skills and who understand how to turn insights into action.
As well as that broad data capability, work should be done in parallel to ensure there’s a clear view of the broader technology system architecture, so that asset registers, GIS, telemetry, maintenance systems, customer information, and finance systems are not isolated systems. Where organisations inherit legacy systems that can’t support integration, there should be a swift, planned move to modern cloud-based platforms.
That won’t happen by accident. It will require a coordinated programme, dedicated funding, and strong leadership willing to treat data and technology as core enabling infrastructure.
Start now with the “no regret” opportunities already in reach
Because of the need for that enabling work on data and technology, realising the benefits from AI is going to be a medium-term project. However, there are still immediate opportunities to use AI, and to build each organisation’s AI posture to be ready for this bigger shift.
There are two broad ways AI could support organisations in the short term.
The first is by improving the productivity and capacity of the teams doing the hard work of establishing the new organisations. Used well, AI can help project teams move faster through common tasks like drafting, summarising, planning, analysing feedback, preparing briefings, and managing large volumes of information. AI adoption done well can support busy teams in practical ways, building their confidence and helping them find additional capacity when the work programme is demanding.
The second use is adopting AI broadly in the corporate environment. Back-office application of AI tools can be broadly transformative. My practical experience at Hutt City Council showed me that even relatively simple use cases can generate real productivity gains across an organisation, if supported by the right training, governance, and leadership. The Council’s programme reclaimed around 44,000 hours a year, with staff using AI to speed up routine tasks and free up capacity for higher-value work. AI also improved drafting and analysis.
Both of those uses of AI – supporting establishment and project teams, and supporting back-office functions – are essentially “no regret” opportunities. They require only off-the-shelf tools that are readily available and easy to buy, like CoPilot, ChatGPT, and Claude. What matters here is strong leadership and a structured adoption programme, with clear expectations for use. This helps teams build AI capability and translate the tools into everyday productivity gains.
The immediate establishment pressures aren’t a reason to wait
It would be tempting to say that the short-term pressures of establishing the new water organisations mean that longer-term transformation has to wait. In reality, the opposite is true. While the new entities are being set up, this is the critical window for making the foundational choices that will make future transformation easier.
Every leadership team should be asking what AI means for its operating model. It will affect how work is designed, how decisions are supported, and how services are delivered. I heard it said recently that it could take 10 years to adopt the AI capability that’s already available today, even if the technology stopped developing tomorrow. That gives a sense of how early we are in this shift, and how much opportunity still sits in front of us.
We have a rare opportunity
The creation of the new water organisations gives New Zealand a rare opportunity to build the digital, data, and technology foundations that are needed to change how water services are planned and managed into the future. In that context, AI can, if used well, help water entities to improve their performance and productivity and to make better decisions.
For chairs, chief executives, and establishment teams, AI adoption shouldn’t be seen as an optional innovation project, but as part of the core work of delivering the water reforms. The opportunity is real, the tools are already available, and there’s no better time to start than now.
