Smart cities are changing how we live, connecting everything from traffic lights to power grids with Internet of Things (IoT) devices [1], [2]. But as our urban spaces become smarter, they face new security challenges. Traditional security methods can't keep up with the scale and complexity of these interconnected systems. This is where Artificial Intelligence (AI) steps in, offering a new layer of protection with its ability to detect threats in real-time, automate responses, and predict future risks [3–5].

Recent developments in AI are proving to be a game changer. For instance, AI-powered systems can now analyze vast amounts of data from traffic patterns and spot anomalies that could signify a threat, all without Predictive analytics: Our system doesn't just react to threats; it predicts them. Analyzing patterns and trends from past data can help anticipate where vulnerabilities might occur next. This foresight allows us to be proactive, fortifying weak spots before they're exploited.

Automated responses: When a threat is detected, time is wasted. That's where automated responses come in. The system can make quick decisions, like isolating a compromised device or tweaking settings to shut down an attack pathway, all without needing a human to step in. This swift action helps keep critical city services running without interruption.

While the benefits of AI-assisted network security in smart cities are substantial, some notable challenges accompany its implementation [12].

Enhanced threat detection: AI's capability to analyze large datasets in real-time increases the accuracy of detecting potential threats before they escalate.

Automated threat response: AI-driven automation ensures a rapid response, minimizing the impact of cyberattacks on essential city services.

Scalability: AI frameworks can evolve with growing numbers of IoT devices and adapt to emerging attack patterns, making them a future-ready solution for urban environments [13].

Privacy concerns: The extensive data collection for AI raises questions about user privacy and data security. Algorithmic bias: Without careful monitoring, AI systems can reflect biases in their training data, potentially leading to unfair or less effective security responses.

Complexity: Integrating AI into security frameworks requires specialized expertise and additional resources for smooth implementation.

AI-driven security solutions are transforming urban management across multiple sectors. Here’s how they’re making a real-world impact.

I. Smart traffic management: AI processes real-time data from cameras and sensors to reduce congestion, dynamically adjusting traffic signals based on current traffic patterns. In practice, this can ease urban gridlock and make commuting faster and safer for city dwellers. By reducing traffic jams and improving flow, cities can reduce pollution and enhance road safety.

II. Energy grid optimization: To manage electricity usage more efficiently, AI-integrated smart meters monitor real-time consumption, flagging unusual patterns that may indicate security risks or inefficiencies. This has translated to notable energy savings and reduced carbon emissions in several cities—particularly in large buildings and public sector facilities, where demand is highest. For example, trials in Suwon, South Korea, led to a 30% increase in energy efficiency and a 35% decrease in emissions from public buildings.

III. Healthcare monitoring: AI supports critical health infrastructure by analyzing data from wearable devices to monitor patient vitals and alert medical personnel during emergencies. This system ensures patient privacy through secure data protocols, guarding sensitive health data against breaches and keeping healthcare systems running smoothly in cities where service demand is growing.

IV. Threat detection and crowd management: AI-powered systems enhance public safety in busy urban areas.

Real-time analysis from cameras and IoT sensors enables AI to detect suspicious objects, recognize faces from watchlists, or even spot behaviors that could indicate criminal intent. This is especially valuable for crowd management during large events or in busy transit hubs, where AI can instantly alert security teams to potential threats.

V. Smart waste and environmental management: Many cities struggle with waste collection inefficiencies. AI-integrated smart bins can monitor fill levels and optimize collection routes, reducing operational costs and emissions from garbage trucks. Additionally, air quality sensors integrated with AI monitor pollution trends, enabling cities to proactively respond to environmental hazards and making urban areas healthier for residents [6].

These applications show how AI fortifies urban infrastructure against cyber and physical threats. By securing critical systems and optimizing resources, AI is helping cities become more resilient, efficient, and safe for residents.

In summary, integrating AI into network security frameworks for smart cities represents a transformative approach that enhances real-time threat detection, automates responses, and maintains secure data flow. With a centralized AI-powered monitoring hub, predictive analytics, and adaptive protocols, cities are better equipped to manage the complexities of urban IoT networks. However, addressing privacy, bias, and system transparency challenges will be essential to fully realize AI's potential in securing these environments. As AI-driven security solutions evolve, they will be instrumental in safeguarding vital city services—such as transportation, healthcare, and energy management—ensuring seamless operation amid the changing landscape of cyber threats [17].

This research received no external funding.

The data used and analyzed in this study are accessible from the corresponding author upon reasonable request.

The authors declare no conflicts of interest. Authors must disclose any personal circumstances or interests that may have an undue influence on the presentation or interpretation of the reported study findings. Funders had no part in the study's design, data collection, analysis, interpretation, manuscript preparation, or decision to publish the findings.