Xylem Vue, the Valencia-based water technology company formed through an alliance between Xylem and Idrica, has set out how agentic artificial intelligence could help utilities strengthen operational resilience, meet regulatory audit requirements and retain the knowledge of retiring staff. The company argues that agentic AI governance must be built into the layer that controls a model’s access to a utility’s operational systems, rather than added after deployment.
Workforce and infrastructure pressures
The water sector faces ageing infrastructure, insufficient investment to renew and expand networks, and a generation of professionals approaching retirement. Xylem Vue cites an estimate by the American Water Works Association that more than 30% of water-sector professionals in the United States will retire by 2030. The US Environmental Protection Agency provides background on water infrastructure sector workforce challenges.
According to the company, these pressures are compounded by climate volatility and growing demand for services that are more resilient, transparent and efficient. It said generative AI is unlocking value in areas whose cross-cutting, unstructured and context-dependent nature had previously placed them beyond the reach of traditional analytics and machine-learning tools.
Agent architectures add a reasoning layer that orchestrates these models, makes them accessible through natural language and extends their reach to processes that previously sat outside the scope of digitalisation. “This is the frontier that is shifting in 2026,” said Jaime Barba, Head of Xylem Vue. “We are not replacing anything that already works; we are opening a second front on challenges that, until now, depended exclusively on human judgement.”
Integrated responses during crises
Xylem Vue’s report, Water Technology Trends: Revolutionising Water Management, states that the challenges ahead are increasingly focused on business resilience. Climate, cyber and infrastructure crises rarely emerge within a single system, and a severe event can simultaneously span SCADA, GIS, CMMS, alarm systems and emergency protocols.
Until now, operators have had to piece this information together mentally under intense time pressure. The company said agent architectures based on MCP allow an operator to ask a question in natural language and receive an integrated response across multiple domains, with the reasoning traced step by step. Contingency plans become dynamic contexts that can be consulted in real time, while the operator retains control over any resulting actions.
Auditability and regulatory reporting
Regulatory reporting obligations are multiplying. In Europe this is driven by implementation of the EU AI Act, while in the United States it comes through EPA and CISA cybersecurity frameworks and several state-level inquiries into the use of AI. Specific water-quality and cybersecurity regulations also apply across the sector.
Xylem Vue said automating regulatory monitoring, report generation and multilingual alerts reduces the administrative burden, but that the main value lies in traceability. “Every step in the agent’s reasoning is recorded: which tools it consulted, which data it used, which recommendation it issued and what a human approved,” said David Torres, AI Lead at Xylem Vue. “This traceability, embedded in the architecture by design rather than added retrospectively, is what turns an AI system into a system that can be audited by a regulator, an internal audit team or a post-incident investigation.”
Retaining operational knowledge
The company links two challenges, the retirement of professionals with decades of tacit knowledge and the need to explain technical decisions to regulators and citizens, to a common technical approach. It said an AI agent connected to a utility’s knowledge base and the sector’s accumulated expertise can propose the action a senior operator would recommend, explain its reasoning and adapt that explanation to the audience.
For junior operators, this is intended to shorten the learning curve and reduce the loss of institutional knowledge. For citizens and regulators, it aims to make decisions explainable and justifiable. Xylem Vue said procedures, operational history and decision-making criteria remain under the utility’s ownership and control, and are neither transferred to the technology provider nor used to train third-party models.
Agentic AI governance built into the architecture
In the model Xylem Vue advocates, governance sits in an intermediary layer between the operator submitting a query, the language model providing the reasoning, and operational systems such as SCADA, asset management and work-order platforms. The company calls this layer a “water-specific harness”. It determines which tools the model invokes and in what order, what the model can access and do, under which conditions, and with what level of human oversight.
In practice, this includes permissions by asset, area and role; a distinction between read-only queries and actions that modify the system; mandatory human validation for consequential operations; a complete audit trail of every interaction; and the ability to reassess the system whenever a prompt, tool or data source changes. The company said this approach is aligned with the requirements of the EU AI Act.
“What matters is not which model the utility chooses, but the water-specific governance layer that determines whether that model is safe to operate on critical infrastructure,” Torres said. “Xylem Vue provides that layer as the technology provider, but the data, operational knowledge, contexts and configuration decisions always remain in the utility’s hands, regardless of the level of AI adoption the organisation chooses.”
H2O Global News has previously reported on Xylem Vue’s view that smart water meters can support Europe’s digital water shift.
Frequently asked questions
What is agentic AI in the water sector?
According to Xylem Vue, agent architectures add a reasoning layer that orchestrates generative AI models, makes them accessible through natural language and extends them to processes previously outside the scope of digitalisation.
How many US water-sector professionals are expected to retire by 2030?
Xylem Vue cites an American Water Works Association estimate that more than 30% of water-sector professionals in the United States will retire by 2030.
What is Xylem Vue’s water-specific harness?
It is an intermediary layer between the operator, the language model and a utility’s operational systems. It controls which tools the model uses, what it can access and do, and the level of human oversight required.
How does Xylem Vue say agentic AI supports regulatory auditability?
Every step in the agent’s reasoning is recorded, including the tools consulted, data used, recommendation issued and what a human approved. The company says this traceability allows the system to be audited by regulators, internal audit teams or post-incident investigations.
Who owns the data used by Xylem Vue’s AI agents?
Xylem Vue says the data, operational knowledge, contexts and configuration decisions remain under the utility’s ownership and control. It says this knowledge is neither transferred to the technology provider nor used to train third-party models.
