Infrastructure & City Planning Weekly AI News
October 27 - November 4, 2025## Cities Getting Smarter with AI Agents
This week brought exciting news about how AI agents are helping cities around the world solve problems that affect millions of people. AI agents are like helpful robots made of computer code. They can watch lots of information from sensors and cameras, learn patterns, and make good decisions without needing a human to tell them what to do each time. These agents are now being used to help cities manage everything from floods to traffic to electricity.
## Stopping Floods Before They Happen
In Jakarta, Indonesia, one of the biggest cities in the world, flooding has always been a serious danger to people who live there. To fix this problem, officials created an AI system that uses agents to predict when floods will happen. The system collects data from rainfall sensors, river gauges, and weather services. When the AI agents predict a flood is coming, they automatically close floodgates, turn on pumps, and send warnings to residents through an app called JAKI. This gives people time to prepare and stay safe. The city is moving from just reacting to floods after they happen to preventing them before they start.
## Making Cities Cleaner and More Efficient
AI agents are also helping cities use energy and water more smartly. In Munich, Germany, the city's utility company uses Microsoft Azure technology with AI to run electric buses better. The AI agents predict how much electricity people will need and help make sure buses are working efficiently. Because of this, about 90% of Munich's electricity already comes from renewable sources like wind and solar power. Over in Singapore, thousands of tiny sensors are placed in water pipes under the city streets. AI agents analyze the data from these sensors in real-time and find leaks automatically. This has cut water waste by 5% and made pipe bursts nearly disappear.
## Building the Power Plants for AI
While cities use AI to solve everyday problems, huge companies are building the infrastructure that powers AI itself. These are called data centers, and they contain millions of computer chips working together. Data centers need enormous amounts of electricity and cooling water. The U.S. Department of Energy announced in July 2025 that it would build AI data centers on four federal properties: Idaho National Laboratory, Oak Ridge Reservation, Paducah Gaseous Diffusion Plant, and Savannah River Site. This shows that AI infrastructure is now a national priority.
Recently in October 2025, the U.S. Air Force took another step by offering to lease land at five military bases to private companies that want to build AI data centers. These bases are Edwards, Davis-Monthan, Arnold, Robins, and McGuire-Dix-Lakehurst. Companies have until November 14 to submit proposals, and winning bids will be announced in January.
## Massive Investments in AI Infrastructure
Private companies are making huge investments too. OpenAI and Oracle announced plans for a $15 billion data center campus in Port Washington, Wisconsin, with partner Vantage Data Centers. This facility will eventually support nearly one gigawatt of computing power—that's enough electricity for about 750,000 homes. Construction will start soon and finish in 2028. The project is part of something called Stargate, which aims to build 10 gigawatts of AI infrastructure across the entire United States. Similar projects are happening in Texas, New Mexico, Ohio, and other states.
## Smart Agents Running the Machines
One particularly exciting development is that agentic AI solutions are being used to run these data centers themselves. NVIDIA, a major technology company, is building an AI Factory Research Center in Virginia with digital twins (which are virtual copies of real systems). AI agents in this system continuously optimize power usage, cooling, and computer workloads. These agents learn over time and make the whole system more efficient and responsive to the power grid. Companies like Tesla, Siemens, and others are contributing technology to make these AI agents smarter.
## The Responsibility Question
Experts are reminding cities to use AI carefully and fairly. Seattle's 2025-2026 AI Plan sets a good example by requiring human oversight, banning harmful uses, and promising transparency. The real work of making smart cities isn't just about having cool technology—it's about making sure the technology actually helps people and doesn't create new problems. Leaders emphasize that coordination between different city departments and long-term political support matter just as much as the technology itself.
## Looking Ahead
As this weekly update shows, AI is moving from experiments to real, large-scale systems that affect how cities work. AI agents that can make smart decisions on their own are becoming central to both city planning and the infrastructure that powers AI itself. The challenge ahead is making sure this happens in ways that are fair, transparent, and truly beneficial to everyone in the community.