Jordan has started using artificial intelligence to tackle something far more physical than chatbots or software: water disappearing from its distribution network.
The country’s Smart Transformation Project for Water and Wastewater Services Management is using AI-powered monitoring to spot leaks, analyze water consumption and identify supply problems earlier. The initiative is funded by the U.S. government and brings together Jordan’s Ministry of Water and Irrigation, the Water Authority of Jordan and local water companies.
Water and Irrigation Minister Raed Abu Saud has described the program as the first project of its kind in Jordan. Early pilot areas have already recorded lower network pressure and reduced water losses, according to officials.
That matters in Jordan more than it would in most places.
The country has just around 61 cubic meters of renewable freshwater available per person annually, according to U.S. government trade data. The commonly cited threshold for absolute water scarcity is 500 cubic meters. Jordan sits well below it.
The AI Is Looking for Problems Humans May Spot Too Late
A leaking underground pipe does not always announce itself. Pressure changes, unusual consumption patterns and mismatches between supply and demand can sit inside network data long before a problem becomes obvious on the street. Jordan’s new system is meant to catch those signals earlier.
The project combines information from supervisory control and data acquisition systems, or SCADA, with smart meters, geographic information systems and enterprise management software. Instead of leaving that information scattered across separate databases and reports, the platform pulls it into a centralized system. AI can then analyze the incoming data and flag unusual behavior.
It can help operators identify potential leaks, forecast demand, decide where maintenance should happen first and react to supply imbalances closer to real time. Authorities also want the system to improve responses to customer complaints. Amman is an obvious priority. More than half of Jordan’s water subscribers are concentrated there.
Jordan Has a Huge Non-Revenue Water Problem
The technology is arriving because Jordan cannot afford to keep losing such a large portion of the water it already produces.
“Non-revenue water” is the industry term for water entering the supply system but never generating revenue. Sometimes it physically leaks from pipes. Sometimes meters are inaccurate. Theft, billing problems and other administrative failures can also make water effectively disappear from the system.
Jordan has historically lost roughly half of its piped water supply to these problems. A United Nations report estimated around 50% of the country’s supply was being lost as non-revenue water, split between physical and administrative losses.
This is not a small efficiency problem tucked away inside a utility company.
When a country has extremely limited freshwater to begin with, recovering water that is already being pumped can sometimes be as important as finding a new source.
Jordan is also pursuing a separate non-revenue water reduction project in southern and southeastern Amman. In March 2026, the Ministry of Investment opened a qualification process for private companies to participate through a performance-based contract with the Water Authority of Jordan and Miyahuna.
AI Could Change How Water Utilities React
There is an interesting shift buried inside this project. Traditional water network management is often reactive. A pipe breaks. Pressure drops dramatically. Someone reports water in the street. Crews investigate. An AI-assisted network can behave differently.
A suspicious flow pattern may be enough to trigger investigation before a leak grows into a major failure. Demand forecasts can also help operators anticipate shortages rather than simply respond once particular areas are already under pressure. Researchers have been exploring exactly this type of system for Jordan.
A 2026 research paper proposed an AI-driven water management framework combining hydraulic modeling, digital twins, SCADA infrastructure and AI agents. In a proof-of-concept based on a 1,164-junction district network in Amman, the system was able to run automated hydraulic simulations, detect flow anomalies and generate network health reports.
Jordan’s government project is broader, but the direction is similar: turn a water network into something that can be continuously measured rather than periodically inspected.
The Next Challenge Is Connecting Everything
Installing AI is probably the easy headline. Making dozens of systems reliably exchange useful data is harder.
The next phase of Jordan’s smart water project is expected to concentrate on linking existing systems, integrating databases and expanding the underlying digital infrastructure. Officials also want to build local expertise in AI, Internet of Things technologies and data analysis so the system does not remain dependent on outside technical support.
That part may determine whether the project becomes a permanent piece of national infrastructure or stays an impressive pilot.
Jordan has no shortage of reasons to make it work. Every leak detected earlier means water that was already extracted, treated and pumped has a better chance of actually reaching somebody’s tap. For one of the driest countries on Earth, that is a fairly practical use of AI.
Sources
- Middle East AI News — “Jordan deploys AI to cut water losses”
- Jordan News Agency (Petra) — “Investment Ministry Launches Qualification Process for Amman Water Loss Reduction Project”
- U.S. International Trade Administration — Jordan Environment and Water Sector
- United Nations Jordan — “Tapped Out: The Costs of Water Stress in Jordan”
- Fasha et al. — “AI-Driven Framework for Adaptive Water Network Management with Proof-of-Concept Implementation: Addressing Non-Revenue Water in Jordan”
