From fragmented information to joined-up decisions

AI adoption across the water sector has accelerated. For many organisations, the question is no longer whether to use AI but how it can create value.

The shift comes at a time when water service providers face growing pressure. Ageing infrastructure, rising investment requirements and increasing expectations from regulators and communities are placing greater demands on decision-makers. Organisations are being asked to justify expenditure, demonstrate confidence in their asset information and explain the rationale behind decisions that may shape water services for decades.

As interest in AI grows, many organisations are discovering that technology alone does not solve information challenges that have existed for years. It can, however, change how we think about them.

Water organisations hold significant knowledge about the networks and assets that support communities every day.

That knowledge lives in asset management systems. Some information sits within inspection reports, technical standards and historic design documentation. Other records are stored in spreadsheets, project files and maintenance histories. Valuable operational knowledge also resides with people who have spent years operating, maintaining and improving assets.

The information is there. Finding and connecting it is often the challenge.

Consider something as common as a renewal program for a critical pipeline. Developing a clear understanding of asset condition may require maintenance records, inspection reports, operational performance data, historic designs and previous investment decisions. Each source contributes to the overall picture, yet those sources rarely live in a single location.

Water organisations do not operate in silos because people are unwilling to collaborate.

They operate in silos because the work is complex. No individual can be across every part of an organisation's assets, operations, customer outcomes, regulatory obligations, environmental responsibilities and investment decisions. Specialisation is how organisations cope with that complexity.

The difficulty emerges when decisions cross those boundaries.

A renewal decision may depend on operational experience held by field teams, asset information held by asset managers, growth forecasts from planners, financial constraints understood by investment teams and service impacts recognised by customer specialists. Each group may hold a valid part of the answer, while no one person sees the whole.

This is where AI introduces a genuinely new possibility.

Its value is not only that it can retrieve information more quickly. It can help assemble, relate and reconcile information across a volume of data, documents and organisational knowledge that would be difficult for any individual to carry. The goal is not to replace specialist expertise. The goal is to allow specialist expertise to contribute to decisions informed by the broader system around it.

Many organisations approach AI with questions about capability. Can it find information faster? Can it analyse large datasets? Can it help teams work more efficiently?

The answer may be yes. Yet AI also exposes the limitations of the information environment around it.

Questions that appear straightforward can become difficult to answer. What is the failure history of this asset? Why was this design approach selected? What guidance applies to this operational process? The information may exist, although it is often distributed across different systems, documents and teams.

In many cases, the challenge is not the technology itself. The challenge is whether the information is connected, accessible and trusted enough to support the effective use of that technology.

In that sense, AI acts as a spotlight. It shines a light on information gaps, inconsistent records and governance challenges that may have remained hidden in the background. That visibility is prompting a broader conversation about organisational readiness.

Most current AI tools answer questions. You ask for information. The system searches, analyses and responds.

The next wave of capability is often described as agentic AI.

Rather than producing a single answer, agents can be assigned specific responsibilities and complete defined tasks using approved information sources, business rules and connected systems. Multiple specialist agents can also work together, each focused on a particular domain, with an orchestration layer bringing the results together.

For a water organisation, this could involve one specialist focusing on asset condition, another evaluating regulatory obligations, another examining investment priorities and another assessing service impacts. Together they can provide a more complete understanding of a decision than any single perspective alone.

In many ways, this mirrors how organisations already operate. Deep specialists contribute expertise in their own domains. The difference is that AI can help connect those domains more effectively.

This does not reduce the need for information foundations. It increases it.

If systems use inconsistent definitions, if records are incomplete or if accountabilities are unclear, AI can amplify those weaknesses. The more connected and autonomous the capability becomes, the more important governance and information quality become.

Public discussion around AI often centres new tools and emerging capabilities. Yet many organisations are finding value by focusing their attention on the foundations beneath the technology.

They understand where critical information is held. They have clear governance that supports how information is managed and maintained. They know who is responsible for it and how it contributes to decision-making across the business.

These activities may attract less attention than a new AI platform, but they create the conditions needed for meaningful outcomes.

For water service providers, the opportunity extends beyond efficiency gains. Better access to trusted information can help people understand evidence more quickly, identify information gaps that influence future investment decisions and access lessons from previous projects without lengthy searches.

The outcome is greater confidence in decision-making. Confidence that information can be trusted. Confidence that assumptions can be traced back to evidence. Confidence that knowledge isn't being lost as organisations change over time.

Infrastructure decisions are ultimately human decisions.

Water organisations operate in a complex environment where investment choices influence affordability, resilience, environmental performance and customer outcomes. These decisions require professional judgement, local knowledge and an understanding of community expectations that technology cannot replace.

AI has an important supporting role, but not because it removes the need for specialists. Its value lies in helping specialist knowledge travel further. It can connect operational experience with asset data, technical standards, investment assumptions and service impacts, making those relationships more visible to the people responsible for decisions. The silos may become more permeable. The specialists do not become less important.

Their role increasingly becomes defining meaning, testing assumptions, recognising exceptions and applying judgement where context matters most.

This is why governance is such a critical part of the conversation. Trust comes from understanding where information originates, how recommendations are generated and who is responsible for the decisions that follow.

For organisations responsible for essential infrastructure, that trust is fundamental.

The next phase of AI adoption in the water sector is unlikely to be defined by the number of tools organisations acquire.

It will be defined by whether those tools improve how decisions are made.

Many of the answers to the sector's challenges already exist within asset records, operational systems, project documentation and the knowledge of experienced people. The opportunity is to make information trustworthy, give it shared meaning and connect it across organisational boundaries.

A practical place to begin is one decision that is difficult, repeated and important.

Identify the information that decision depends on. Identify the people whose expertise contributes to it. Agree who remains accountable for the outcome. Then explore whether AI can help bring the evidence together in a way that is more transparent, consistent and useful.

AI will not remove the need for good information management. It will not remove the need for people.

Its more compelling promise is that it may allow organisations to use both more effectively, helping specialist knowledge contribute to decisions informed by the whole system rather than a single part of it.

That is where strong information foundations become valuable. And that is where AI can create lasting impact.