Smart Cities Secured: Utilizing AI Firewalls for Sustainable Urban Environments

AI firewalls are essential in enhancing the resilience of smart cities, which are increasingly dependent on interconnected technologies such as the Internet of Things (IoT) and cloud computing. These technologies provide critical urban infrastructures such as energy management, transportation, health care, and public safety. However, the growing interdependence of these systems makes smart cities more vulnerable to cyber threats. As traditional firewalls struggle to maintain the complexity and complexity of evolving cyber threats, AI-powered firewalls provide a more adaptive and proactive solution. AI firewalls use ML and deep learning (DL) algorithms to monitor real-time network traffic, detect anomalies, and immediately respond to potential cyberattacks. This allows them to provide dynamic protection for smart city infrastructures, even as new devices and systems are continuously integrated into the environment. By detecting and mitigating risks before they cause considerable damage, AI firewalls can maintain the continuous operation of vital services such as energy grids, traffic systems, and health care facilities. In doing so, they contribute to the reliability, security, and overall resilience of smart cities, preventing disturbances and ensuring uninterrupted service delivery. Additionally, AI firewalls are crucial in enhancing public trust by safeguarding sensitive citizen data, which is crucial for the successful adoption of smart city initiatives. Their ability to balance security and privacy encourages greater citizen participation and promotes the sustainable development of smart cities. In summary, AI firewalls are essential not only to protect smart city infrastructures but also to ensure their resilience against cyber threats, creating a secure and sustainable urban environment for residents.

AI firewalls are crucial in boosting the resilience of smart cities by providing advanced cybersecurity mechanisms that address the unique challenges faced by interconnected urban systems. Smart cities, led by IoT devices and cloud technologies, utilize seamless data flow and communication between critical infrastructures, such as energy management, health care, transportation, and public safety. As these systems become increasingly complex, they also become more vulnerable to cyber threats such as data breaches, system disruptions, and denial-of-service attacks. AI firewalls provide a dynamic and adaptive approach to these issues by utilizing ML and DL algorithms to analyze traffic patterns, detect anomalies, and respond autonomously to emerging threats. Like traditional firewalls, which are static and based on predefined rules, AI firewalls can rapidly identify new attack vectors and mitigate them in real time, ensuring the uninterrupted functioning of essential services. This ability to recognize and respond to threats enhances the security and reliability of smart city infrastructures, which is crucial for maintaining public safety, privacy, and trust. By preventing service disruptions and protecting sensitive data, AI firewalls enhance the overall resilience of smart cities, making them more resilient against cyberattacks and operational failures.

Integrating cybersecurity, sustainability, and smart city operations is crucial for creating resilient, forward-thinking urban environments. As smart cities depend heavily on interconnected systems to manage essential infrastructures, including energy, transportation, health care, and public safety, the increasing use of IoT devices and cloud technologies offers both improved efficiency and greater vulnerability to cyber threats. In this context, cybersecurity becomes a crucial component, ensuring that these systems operate safely while also promoting the sustainable objectives of smart city initiatives. AI firewalls are positioned at the intersection of cybersecurity and sustainability, offering dynamic, adaptive security solutions that safeguard vital urban operations.

Smart cities are designed to be constructed on interconnected networks that ensure the continual operation of urban services. Cybersecurity is a vital component of protecting these networks from attacks that could disrupt services and systems. AI firewalls enhance this security by detecting and responding to cyber threats in real-time, enabling the protection of key infrastructures such as energy grids, public safety networks, and transportation systems. By safeguarding these systems, AI firewalls prevent potential disruptions that would otherwise decrease efficiency and carbon emissions. In addition, by maintaining the privacy and security of citizen data, AI firewalls create trust, promoting public participation in sustainability efforts such as waste management and resource optimization. They also ensure that as new devices are added to the network, these enhancements can be seamlessly integrated into the security framework without compromising system integrity, thus fostering sustainable urban development.

AI firewalls are crucial in enhancing the resilience of smart cities by providing robust cybersecurity solutions tailored to address the unique challenges faced by the ever-expanding and interconnected urban systems. As smart cities increasingly utilize IoT devices and cloud technologies, real-time data management and seamless communication between systems such as energy, transportation, health care, and public safety become crucial. However, the increasing complexity of these systems presents them with a variety of cyber threats that can undermine their integrity and disrupt urban operations. AI firewalls, utilizing ML and DL algorithms, actively monitor network traffic to detect and mitigate these threats, ensuring the uninterrupted functioning of critical urban infrastructures.

Unlike traditional static firewalls, AI firewalls provide dynamic protection by constantly learning and adapting to emerging threats, making them essential in an environment where attack methods evolve rapidly. This adaptability helps to safeguard basic services such as health care, energy management, and traffic control, all of which depend on continuous, secure data flow. The deployment of AI firewalls enhances the security of these systems and contributes to the sustainability of intelligent cities. By preventing service disruptions and ensuring that resources such as energy and data are efficiently managed, AI firewalls play a fundamental role in protecting both the digital and physical infrastructure of cities, allowing them to withstand and recover from cyber incidents or system failures, thus fostering resilience and sustainability in urban environments.

The rapid expansion of IoT devices in smart cities significantly improves operational efficiency but also exposes significant cybersecurity vulnerabilities. Many IoT devices possess limited security features, such as weak encryption, poor authentication protocols, and frequent firmware updates, making them susceptible to cyberattacks. As these devices operate in highly interconnected environments, they create expansive attack surfaces that cybercriminals can exploit to gain unauthorized access or disrupt services. AI firewalls are crucial in addressing these vulnerabilities by utilizing advanced technologies such as ML and DL. Unlike traditional firewalls, which rely on static, predefined rules, AI firewalls analyze network traffic continuously, detecting anomalies and potential real-time threats. This dynamic, adaptive approach is particularly beneficial for ensuring IoT ecosystems, where the volume and complexity of data flows require intelligent, automated security solutions. Key functionalities of AI firewalls in securing IoT devices include anomaly detection, where they monitor baseline behaviors of devices and identify deviations that could signal malicious activity, such as unusual data patterns or unauthorized access attempts; predictive threat analysis, where historical data help predict emerging attack vectors, enabling proactive measures to thwart potential threats; automated incident response, allowing the firewall to autonomously isolate compromised devices, block malicious traffic, and restore normal operations without manual intervention; and self-learning capabilities, enabling the firewall to update its detection models continually to adapt to new attack techniques. AI firewalls also enhance the security and interoperability of IoT networks, ensuring consistent security across various domains such as transportation, health care, energy management, and public safety. Edge AI firewalls, which process data closer to the source, reduce latency and improve real-time threat detection, a critical feature for applications such as stand-alone vehicles and emergency response systems. Furthermore, AI firewalls address privacy concerns by utilizing federated learning and differential privacy, ensuring secure training of AI models on distributed data without exposing sensitive information. In conclusion, the deployment of AI firewalls is essential for ensuring the interconnected IoT networks of smart cities. By offering adaptive, scalable, and intelligent security, AI firewalls protect critical infrastructure, safeguard citizen data, and ensure the seamless operation and sustainability of smart urban environments, highlighting the need for intelligent cybersecurity in an increasingly connected world.

AI firewalls are a significant advancement in securing the IoT devices in smart cities. Given the increasing reliance on interconnected devices that form the backbone of smart city ecosystems, securing these IoT devices becomes essential to ensuring public safety, privacy, and the functionality of critical urban services. Traditional firewalls, which utilize pre-set rules and signatures to detect malicious activities, are unable to handle the dynamic and large-scale nature of IoT networks. In contrast, AI-powered firewalls utilize ML and DL algorithms to detect real-time threats by analyzing vast amounts of data traffic and detecting patterns indicative of potential breaches or attacks. These intelligent firewalls learn from new data, adapting to the evolving cyber threat in smart cities. In the context of IoT, AI firewalls can detect anomalies across a network of diverse, often resource-constrained devices, providing enhanced protection against cyberattacks such as distributed denial of service (DDoS), man-in-the-middle attacks, and device manipulation. The ability to predict and respond to threats independently allows for faster mitigation and less disruption to essential services such as traffic management, health care, energy distribution, and public safety.

The role of AI firewalls extends beyond threat detection. They enable autonomous responses to cyber incidents, enhancing real-time security across the IoT networks in smart cities. For instance, when an anomaly or potential threat is detected, AI firewalls can isolate the affected device or network segment without manual intervention, preventing further attack spread. This quick action is crucial in high-stakes environments where delays can lead to considerable damage, such as in critical infrastructure systems. Additionally, AI firewalls provide robust and adaptive security, ensuring that even as new IoT devices are added to the network, the security system continues to provide robust protection without requiring extensive manual reconfiguration. This proactive, dynamic approach to cybersecurity is crucial for the ongoing protection of IoT devices in smart cities, where networks are constantly changing, and cyber threats are constantly evolving.

Table 1 compares traditional firewalls to AI firewalls in the context of securing IoT devices in smart cities. While traditional firewalls rely on static, predefined rules and signatures for threat detection, AI firewalls utilize ML and DL algorithms to detect both known and unknown threats based on data patterns, offering more adaptive and scalable security. AI firewalls provide faster, stand-alone real-time responses, reducing reliance on manual intervention, while traditional firewalls may experience slower response times. AI firewalls also excel in anomaly detection and resource efficiency, learning from new data, while traditional firewalls struggle with scalability, and require frequent updates. This enhances AI firewalls for dynamic, large-scale IoT networks in smart cities, providing robust, long-term protection against evolving cyber threats.

The study on AI firewalls for smart cities emphasizes their crucial role in enhancing urban cybersecurity and fostering sustainability in interconnected digital infrastructures. As cities become increasingly dependent on IoT, AI, and ML technologies, they are exposed to heightened cybersecurity risks, particularly due to the interconnectedness of critical services such as transportation, energy, and public safety. AI firewalls provide a significant improvement over traditional security measures by leveraging ML and DL algorithms to automatically detect, analyze, and respond to cyber threats in real time. This proactive approach enables them to continuously learn and adapt, ensuring robust protection against evolving attack patterns. Key functionalities of AI firewalls include anomaly detection, predictive threat analysis, automated incident response, and self-learning adaptability, which enhances smart cities’ resilience. The integration of AI firewalls strengthens cybersecurity and contributes to the sustainability of urban systems by optimizing energy usage, protecting data privacy, and ensuring the continuity of essential services. Despite the promising benefits, challenges remain, such as excessive deployment cost, data privacy concerns, and scalability issues. Future research is focused on emerging trends such as federated learning, edge AI firewalls, and explainable AI (XAI) to address these limitations and enhance the effectiveness of AI firewalls. Furthermore, AI firewalls are a crucial component of the digital transformation of smart cities, ensuring security, resilience, and public trust while advancing sustainability and efficiency in urban areas.

AI firewall technologies play a pivotal role in enhancing public trust in smart cities by addressing critical challenges associated with data protection and privacy. In smart cities, where vast amounts of sensitive personal and operational data are transmitted through interconnected systems, safeguarding this information is essential to maintaining citizen’s confidence. AI firewalls, powered by advanced ML and DL algorithms, provide real-time anomaly detection, predictive threat analysis, and automated incident responses, ensuring that cyber threats are identified and mitigated before they can compromise data integrity. Their ability to learn and adapt to evolving attack patterns enhances the robustness of cybersecurity frameworks, making them indispensable in urban areas. Moreover, these technologies incorporate privacy-preserving mechanisms such as federated learning, which enables AI models to learn from distributed data without requiring centralized collection, and differential privacy, which protects individual data points from exposure while retaining analytical value. This ensures that citizen’s personal information remains secure even as smart city systems become more complex and data-intensive. By proactively addressing potential vulnerabilities, AI firewalls not only secure critical infrastructure but also demonstrate a commitment to transparency and ethical data handling. This enhances public confidence in the safety and reliability of digital urban systems, encouraging greater citizen participation and fostering trust in the technological framework of smart cities. Furthermore, AI firewalls are essential elements in the digital transformation of urban landscapes, ensuring both cybersecurity and the trust necessary for the sustainable development of smart cities.

The integration of AI firewalls in smart city ecosystems contributes significantly to the sustainability of energy consumption and reduces environmental impact. These advanced systems enhance cybersecurity and play a pivotal role in ensuring efficient resource utilization within interconnected urban networks. AI firewalls can effectively monitor and manage energy flows by analyzing data traffic patterns, identifying inefficiencies, and mitigating vulnerabilities that could lead to energy wastage or system disruptions. For example, by preventing cyberattacks that could compromise critical infrastructure such as smart grids, AI firewalls help maintain uninterrupted and efficient energy distribution, reducing the carbon footprint associated with operational downtime or resource overuse. Additionally, their ability to prioritize energy-efficient protocols for IoT devices and infrastructure ensures that only necessary computational resources are utilized, thereby reducing overall power consumption. These systems also contribute to sustainable practices by supporting predictive maintenance, where potential faults in power systems are detected early, preventing failures that could otherwise lead to excessive energy use or costly repairs. Furthermore, AI firewalls enable municipalities to comply with environmental regulations by safeguarding data integrity and enhancing transparency in energy management practices, further enhancing public trust in smart city initiatives. As these firewalls evolve, their contribution to green urban living will expand, incorporating advanced ML algorithms to enhance the efficiency of renewable energy systems, thus ensuring that smart cities not only remain secure but also environmentally sustainable.

The intricate relationships between cybersecurity, sustainability, and smart city operations provide a crucial basis for establishing and sustaining contemporary urban landscapes. Cybersecurity plays a vital role in smart cities by safeguarding interconnected systems, including essential infrastructure, IoT devices, and residents’ information. These systems are exposed to attacks that could impair vital services such as health care, transportation, and utilities, resulting in dire consequences for urban functionality and inhabitants. Given smart cities’ reliance on digital technologies, safeguarding these systems from cyber threats is essential for maintaining operational stability and ensuring the safety and resilience of urban infrastructure. In the context of smart cities, sustainability is increasingly concerned with cybersecurity, as the protection of digital infrastructure ensures the long-term viability of urban systems and their ability to adapt to environmental and technological changes. For instance, AI firewalls shield data and networks from cyberattacks and contribute to energy optimization by efficiently monitoring and managing resource usage. These technologies reduce the environmental impact of urban operations while maintaining a secure and secure digital ecosystem. The integration of sustainable practices with cybersecurity enhances the ability of smart cities to support sustainable urban life by ensuring that critical services can continue uninterrupted, even when faced with emerging threats or system vulnerabilities. Furthermore, the interplay between cybersecurity, sustainability, and smart city operations generates a synergistic effect, where secure, resilient infrastructure enables smart cities to flourish sustainably. AI-powered security solutions such as firewalls are instrumental in this dynamic, enhancing both cybersecurity and operational efficiency while contributing to the city’s wider sustainability objectives. The convergence of these elements ensures that smart cities can continue to expand, evolve, and address emerging challenges without compromising the security of their systems or the environmental sustainability of the urban ecosystem.

The integration of AI-driven firewall systems in smart cities presents significant challenges as well as promising opportunities. One substantial obstacle is adapting AI-driven firewall system to the diverse and comprehensive systems that comprise the smart cities landscape, such as traffic management systems, health care, utilities, smart home ecosystems and public services. These systems require interoperable and interconnected frameworks to efficiently manage the variant network traffic. Moreover, the increased penetration of IoT devices increases the difficulty of ensuring the safety of these networks against digital threats, necessitating real-time threat detection and prompt response capabilities. The issue of citizen data privacy concerns is one more problem, as the extensive use of large data in urban areas must balance analytical requirements with the protection of citizen data. Techniques such as differential data privacy and federated learning are essential to prevent these cyber risks. Furthermore, the automation of incident response and recovery processes is dynamic to ensure minimal downtime and resilient sturdiness, but it requires intricate regenerative processes to be integrated into AI-driven firewall system.

As urban centers become smarter cities, they become increasingly reliant on interlinked digital systems, growing their vulnerability to cyber threats. Adopting sophisticated technologies such as IoT, AI, and ML for urban management presents cybersecurity risks. The consistent nature of these systems means that a security violation in one area can cause considerable damage to other critical services, including transportation systems, energy management, and health care systems. Smart cities often use IoT devices that are susceptible to cyberattacks due to their limited processing capabilities, flawed encryption, and limited defense features. This leaves smart city areas exposed to unauthorized access, data breaches, and system disruptions. Additionally, scaling traditional classical security systems to meet the expanding needs and complexities of the smart city landscape poses a significant obstacle. As these urban landscapes grow, the number of IoT devices, applications, and data sources necessitates more adaptive, real-time application, and agile defense solutions.

However, AI-driven firewall systems provide immense opportunities to enhance the cybersecurity and sustainability of urban digital landscapes. By utilizing advanced ML models, these systems can transition from reactive to proactive security methods, predicting and neutralizing cyber threats before they are implemented. The incorporation of edge computing reduces latency and enhances threat detection’s real-time competences, making the AI firewall system essential as IoT adoption increases. Furthermore, the scalability and interoperability of AI security systems provide a solid foundation for improving the urban landscape without compromising safety. Beyond cybersecurity, AI-driven firewall systems contribute to energy-efficient operations, sustaining smart cities’ infrastructures. Their ability to promote public trust in digital systems ensures the successful development of technology-driven urban innovations. As AI firewalls evolve, their collaboration with urban governance and policymakers will ensure ethical and privacy-conscious deployment, strengthening their role as a key component of secure and resilient smart cities.

AI-powered firewalls are instrumental in addressing these cybersecurity challenges. In contrast to conventional firewalls, AI-driven alternatives use ML and DL algorithms to examine data patterns, identify anomalies, and independently respond to cyber threats. These systems continuously learn from the data they process, adapting to emerging threats and evolving attacks without manual intervention. This self-learning capability makes AI firewalls particularly effective for the dynamic environment of smart cities. Their ability to perform real-time anomaly detection and predictive threat analysis allows them to identify potential vulnerabilities before they are exploited. Furthermore, AI firewalls can automate incident responses, such as isolating compromised devices or blocking unauthorized access, thus minimizing the impact of attacks. By safeguarding IoT networks, AI firewalls ensure the uninterrupted operation of essential services such as energy grids, waste management, and public transportation, contributing to the sustainability and sustainability of smart city infrastructures. Their scalability and adaptability make them ideal for the complex, evolving nature of smart cities, providing ongoing protection as new devices and systems are implemented. AI firewalls also safeguard data privacy, fostering public reliance by preventing unauthorized access to sensitive personal citizen data information. AI firewall system improves cybersecurity and facilitates sustainable development and resilient urban landscapes, ensuring the long-term viability and integrity of smart city systems.

Municipalities are facing numerous challenges when deploying AI-driven firewalls for smart city landscapes, resulting from financial, technical, organizational, and regulatory challenges. One of the most significant obstacles is the excessive cost of implementation, as AI-powered cybersecurity systems require substantial investments in advanced hardware, sophisticated software, and skilled experts. Budget constraints often make it difficult for municipalities to prioritize cybersecurity over other essential public reliance services. Furthermore, integrating AI firewall systems into existing smart city systems is a challenging task, given the diversity of IoT devices, security systems, and heterogeneous network architectures. It requires extensive customization and the ability to ensure seamless integration across these IoT platforms. Data privacy concerns further complicated deployment, as AI-driven firewall systems utilize vast amounts of data to detect cyber threats, raising questions about the ethical collection, storage, and processing of sensitive citizen personal data. Municipality areas must comply with strict data privacy regulations such as GDPR or CCPA, ensuring public reliance is not compromised.

Furthermore, the lack of skilled personnel in AI, network security, and IoT systems poses a significant obstacle. Municipalities often struggle to attract, retain valuable individuals due to competition from the private sector, limiting their ability to manage and maintain advanced security systems efficiently. The ever-evolving cyber threat landscape compounds this obstacle, as opponents increasingly adopt AI-based techniques to break cyber defenses. Keeping AI firewall systems adaptive and resilient to emerging cyber threats requires ongoing monitoring and updates, which require additional resources and expertise. Moreover, interoperability issues further complicate deployment, as many urban systems operate on proprietary protocols or incompatible standards, making it challenging to achieve streamlined integration without significant effort and expense. Smaller municipalities often lack the financial and technical resources to adopt AI-driven firewalls, creating a disparity in cybersecurity preparedness between urban centers and impoverished regions.

Additionally, regulatory and security policy ambiguities pose an obstacle, as many areas lack clear guidelines for the use of AI-driven cybersecurity in the landscape. Municipalities must navigate these uncertainties, which can delay decision-making and deployment efforts in the network. The resistance to change among stakeholders, including citizens and internal teams, also presents challenges, as modern AI-based technologies are often encountered with skepticism or concern over their effectiveness and implications. Building reliance and demonstrating the value of AI-driven firewalls can be time-intensive and requires extensive communication and stakeholder engagement. Finally, AI-driven firewall systems require ongoing maintenance, updates, and cyber threat intelligence to remain effective, necessitating a long-term commitment of system resources for operational costs, which can potentially strain already limited municipal funds. These obstacles require a collaborative effort among government agencies, private sector partners, and academic institutions. Public–private collaboration, targeted funding initiative system, and knowledge-sharing platforms can collaborate with municipalities to address these obstacles, empowering the successful implementation of AI-driven firewall systems to secure interconnected smart urban environments.

Smart cities represent the integration of information and communication technologies (ICTs) into urban areas to enhance landscapes, optimize resource usage, and enhance the social welfare for residents. This digital transformation affects many aspects of the urban management system, including the transportation system, energy distribution, waste management, health care system, and public reliance (Chourabi et al., 2012; Zhang & Lee, 2020).

However, as urban systems become increasingly interlinked through IoT devices and cloud computing-based applications, they also become more vulnerable to cybersecurity threats. The interconnected nature of the smart city landscape means that any breach in one part of the system can jeopardize critical services, posing risks to citizens’ privacy, public reliance, and infrastructure integrity (Yin et al., 2020). For instance, a cyberattack on a smart grid could disrupt energy supplies, impacting network traffic signals, health care facilities, and public protection activities (Brown et al., 2023).

IoT devices play a significant role in the smart city landscape, supporting the monitoring and control of various smart city systems. However, many IoT-enabled devices are inherently vulnerable due to limited processing power, which restricts their protection mechanisms (Zhu et al., 2021). Additionally, these devices often rely on minimal cryptographic encryption and may lack strong authentication protocols, making them strategic priorities for cybercriminals. Attackers can exploit these vulnerabilities to gain unauthorized access, misappropriate data, and disrupt operations across the smart city landscape.

The unique security obstacles posed by IoT networks in smart cities highlight the need for innovative digital protection solutions capable of managing complex, large-scale data flows and recognizing potential cyber threats in real-time applications (Patel & Sinha, 2022). Traditional classical firewalls are insufficient to protect against sophisticated attacks due to their inability to adapt and learn from evolving attack patterns, which is why researchers have turned to AI firewalls as a potential solution (Xu & Park, 2024).

AI firewalls have emerged as a transformative approach to addressing cybersecurity concerns in smart cities. Unlike traditional classical firewall systems, which rely on pre-defined access rules and manual updates, AI firewall systems utilize ML and DL algorithms to analyze data patterns, detect threat anomalies, and respond to cyber threats autonomously (Rahman & Ali, 2022). AI-driven firewalls can continuously learn and adapt from data collected across urban systems, enabling them to recognize new threats and reduce response times significantly (Chen et al., 2024).

AI-driven firewalls can manage high volumes of data from various smart city applications in real-time systems, identifying and mitigating security risks without manual or human intervention. Their ability to scale and adapt is particularly valuable in dynamic urban landscapes, where new IoT devices and data streams are continually expanding, and cyber threats evolve rapidly (Martin & Lee, 2023).

The key functionalities of AI-driven firewalls in smart cities include:

Anomaly detection: AI algorithms in firewalls analyze network traffic patterns and identify deviations indicative of potential cyber threats. This real-time system anomaly detection is crucial for dynamic landscapes such as smart cities, where multiple devices and applications operate simultaneously (Singh & Zhang, 2022).

AI firewall systems play a dual role in promoting sustainability within smart cities by enhancing cybersecurity and resource efficiency. Securing IoT networks prevents the disruption of essential services such as energy grids, waste management systems, and public transportation, which are critical for sustainable smart city operations (Hassan et al., 2024). AI firewall system also improves resource efficacy by enabling an intelligent energy management system, reducing the need for manual monitoring, and minimizing the carbon footprint of smart city services.

From the perspective of sustainable urban development, AI-driven firewalls contribute to:

Efficient energy use: By managing security in smart grids and optimizing data flow, the AI firewall system can reduce the energy consumption of IoT devices and data centers, aligning cybersecurity efforts with sustainable resource management (Chen & Huang, 2021).

Despite the advantages of AI-driven firewalls, their implementation in smart cities presents several obstacles:

Cost of deployment: The incorporation of AI security systems requires significant investment in both hardware and software, which may be prohibitive for smaller municipalities. Moreover, the prohibitive cost of skilled personnel for maintenance and updates can stress municipal fiscal allocations (Gomez et al., 2023).

As AI firewall technology continues to evolve, researchers are exploring ways to overcome present constraints and improve the effectiveness of AI-based cybersecurity solutions. Some promising areas for future research direction include:

Federated learning: Federated learning allows AI firewall systems to learn from distributed data sources without compromising data privacy. This approach can enhance data security in smart cities by reducing the need for centralized data collection (Jung & Lee, 2024).

The rapid deployment of IoT devices and cloud computing technologies in urban landscapes has led to the development of smart cities, which aim to improve the quality of life, optimize system resources, and improve urban management systems. However, this smart transformation also brings significant challenges to the cybersecurity system. As cities become more interconnected, traditional classical firewalls, which rely on preconfigured access rules and manual updates, are increasingly unable to manage the escalating intricacies of cyber threats. This is where AI-powered firewall systems provide an advanced and adaptive solution by utilizing ML and DL algorithms. AI-driven firewalls are designed to handle large-scale and dynamic data flows characterizing the smart cities’ landscape. Unlike traditional classical firewalls, they can analyze network traffic in the real-time system, detect threat anomalies, predict potential cyber threats, and autonomously respond to cyberattacks. Their continuous learning and adaptation to emergent attack patterns are crucial for preventing the cybersecurity risks posed by the intricate IoT systems in the smart cities’ landscape. Key features of the AI firewall system include anomaly detection, predictive cyber threat analysis, automated incident response, and self-learning adaptability, all of which are crucial in dynamic landscapes. They secure vital infrastructures, such as power grids, health care systems, and traffic networks, by monitoring data flows and responding to potential cyber threats. For instance, AI firewall systems can protect smart grids by detecting anomalies in power distribution systems, secure health data from cyberattacks, and help protect traffic management systems that rely on real-time data analysis for municipal mobility. By enhancing security and data privacy, AI-driven firewalls can foster trust in smart city technologies, ensuring citizen’s data remains secure and encouraging participation in urban initiatives. Despite their advantages, implementing the AI firewall system in smart cities poses challenges, such as high deployment costs, citizen data privacy concerns, and issues with scalability and interoperability. The costs of AI-driven firewalls include investments in hardware and software and the need for skilled staff or experts to maintain and update the system. Moreover, AI firewalls use large datasets for ML models, causing concerns about how to balance privacy with security. Additionally, as the number of IoT devices increases, the AI firewall system must scale to accommodate these changes and achieving interoperability across multiple IoT systems relies on consistent communication protocols, which are often lacking. To overcome these obstacles, emerging technologies such as federated learning, edge AI-driven firewalls, and XAI are being investigated. Federated learning enables the AI firewall system to train on decentralized data, enhancing data privacy while enhancing cyber threat detection. Edge AI-driven firewalls reduce latency by processing data closer to IoT devices, which is particularly beneficial for real-time systems such as autonomous vehicles and traffic management systems. XAI provides transparency in AI-driven firewall decision-making, helping stakeholders understand and trust the system’s actions while ensuring compliance with data protection policies. In conclusion, AI-driven firewalls are a transformative approach for enhancing cybersecurity in the smart cities landscape, providing advanced, adaptive cryptographic security mechanisms to address the growing complexity of urban networks. While obstacles related to cost, data privacy, scalability, and interoperability persist, ongoing research and technological evolution have the potential to significantly contribute to the cybersecurity, sustainability, and resilience of smart cities’ system, ensuring a safe and effective urban landscape for all residents.

Despite the growing interest in AI-driven firewalls for smart cities, there are numerous important flaws in the existing literature. Many research studies focus on specific sectors, such as energy grids or transportation, but not on the incorporation of AI-driven firewalls across the interlinked systems of a smart city. This limitation creates a gap in understanding how the AI firewall system can manage the complexity and scale of urban landscapes, where various IoT devices interact and depend on each other. Another substantial gap is the scalability and adaptability of AI-driven firewalls in dynamic smart city networks. While AI firewall systems determine promise in enhancing security, there is inadequate research on their ability to scale the increasing number of IoT systems and data sources, or how they can adapt to rapidly evolving cyber threats. Furthermore, citizen data privacy and ethical considerations remain inadequately examined, as AI firewall systems require access to vast amounts of data for training ML techniques, raising concerns about the balance between security and data privacy. Current research studies provide limited guidance on how to implement AI-driven firewalls in accordance with privacy laws and ethical principles, particularly when handling sensitive citizen personal data. The excessive cost of deploying the AI firewall system is another issue not adequately addressed in the literature review, particularly for smaller municipal areas with limited budgets. More research is needed to evaluate the cost-effectiveness and affordability of AI-driven firewalls in real-world urban landscapes. Moreover, despite their focus on safeguarding IoT networks, there is little research on the interoperability of AI firewall systems across the various IoT systems commonly found in smart cities’ systems. The lack of standardized protocols complicates the integration and efficacy of security across multiple IoT devices and platforms. Finally, while AI-driven firewalls are praised for their real-time system detection and prompt response capabilities, there is a scarcity of empirical research studies examining their actual performance in high-data volume, fast-paced environments of urban, highlighting the need for more research to assess their efficiency in real-world implementations.

The incorporation of ICTs into urban systems has led to the emergence of digital cities, renovating how infrastructures, resources, and services are managed in interlinked networks. Leveraging technologies such as IoT and AI, this digital shift promises to lift urban efficacy and improve occupant’s quality of life. However, the interconnectivity that drives smart cities puts them at a significant risk in terms of cybersecurity, as a breach in one system could easily affect others, threatening public reliance and privacy. The IoT devices, at the core of a smart city’s landscape, are susceptible in many ways because they lack some cybersecurity features due to their rudimentary nature, leading to a need for adaptive and emergent cybersecurity solutions. Traditional classical firewalls can do little to counter such cyberattacks, so AI firewalls have become the next frontier of interest among researchers, those that use ML to autonomously cyber detect, adapt, and respond to evolving cyber threats [6, 7].

The intelligence firewall systems can offer functionalities critical to the security of a smart city, such as real-time anomaly cyber threat detection and predictive threat analysis, with automated prompt response. Being an autonomous detection and mitigation tool of cyber threats, AI-driven firewalls can be vital for supporting the continuity of essential smart city services without interruptions, hence further securing infrastructure and social reliability by ensuring the urban operating systems remain in a safe state. AI firewall systems also support sustainable development in urbanism by optimizing resource usage while keeping the systems safe from sustaining urban operations. They contribute toward energy efficiency, citizen data privacy, and resilient infrastructure, issues important for sustainable urban development [8, 9].

Future developments seem to deal with these issues via innovations in federated learning, edge AI firewalls, and XAI. All such innovations promise enhanced data security, faster response times, and greater transparency. As long as AI firewall technology continues improving, all such breakthroughs will be instrumental in ensuring safe and sustainable smart city development.

Integrating AI-driven firewalls into the existing digital infrastructure of smart cities requires a meticulous classical approach to ensure compatibility, performance, and scalability across diverse digital technologies. Smart cities utilize a wide range of security systems, including IoT devices, edge computing systems, cloud platforms, and traditional classical IT networks, creating a complex environment for deploying firewall systems. AI-powered firewall system must be designed to work with these heterogeneous systems, supporting standard communication protocols and providing low-latency processing to ensure real-time cyber threat detection without disrupting crucial services such as transportation or energy grids. Key prerequisites for successful incorporation include a robust network topology capable of high-bandwidth communication between IoT devices and the cryptographic security system, sufficient computational resources for AI-driven model training and inference, and standardized data formats that enhance system cohesion. However, the obstacles arise when dealing with security systems that lack the necessary interfaces or processing capabilities for AI-powered cyber threat analysis. To address these cyber issues, network components may need to be upgraded, edge computing nodes deployed, or interoperability platforms implemented to resolve the compatibility issues. Moreover, a centralized cyber threat intelligence framework is essential for aggregating and analyzing data from distributed secure systems, ensuring that security access policies are consistent across various components and cyber vendors. Federated learning techniques are often used to maintain data privacy and regulatory prompt compliance, particularly when integrating firewall systems in sensitive areas such as health care systems or financial networks. While these technical challenges are substantial, proactive planning, infrastructure upgrades, and adherence to industry standards make it possible to seamlessly integrate AI-driven firewalls into smart city landscapes, enhancing cybersecurity while fostering innovation and growth within smart city ecosystems.

In Table 3, represents the comparative research analysis, which reveals that previous research focused on specific vulnerabilities in urban systems or initial AI implementations. The current work advances the field by presenting AI-driven firewall system as comprehensive solutions, enhancing cybersecurity while promoting sustainable urban landscapes. It fills critical literature gaps by proposing scalable, adaptive, and privacy-compliant methods, positioning AI firewall systems as indispensable tools for smart cities’ future.

The first phase in implementing AI firewalls within smart cities is the significant analysis and planning phase, where one defines the main objectives that the AI firewall will for the underlying smart city infrastructure. This will include the protection of the city’s urban IoT network, safeguarding the connection of devices against cyber threats, protecting citizen data and personally identifiable information collected through services, and service continuity, ensuring the continuous running of the essential services despite a potential cyber incident. After that, risk assessment for cybersecurity about important city systems such as traffic management, utilities, health care, and public safety is done to recognize vulnerable areas and define the scope of AI firewall components. A relevant case study is the Barcelona Smart City Initiative, a 2021 city-wide risk assessment focused on critical sectors such as energy and transportation, creating a basis for implementing targeted AI firewall solutions to improve cybersecurity for smart infrastructure within the city.

In Phase 2, data collection and preprocessing, distributed IoT sensors and edge computing devices in the digital city collect a stream of several types of data, including network traffic, access logs, and sensor data from different urban applications. Collected data must then go through a preprocessing stage before being used in training AI models. This stage involves anonymizing penetrating data to preserve privacy and filling in missing values to guarantee data integrity. The data is then divided into training, validation, and testing datasets, which helps create accurate and unbiased intelligence models.

In the third phase of development, model progress and training, more refined ML algorithms, including K-means clustering and deep neural networks, would cultivate an anomaly threat detection model that could identify network deviations from regular behavior. This model is then trained in history data where typical user patterns are learned to help detect unusual or suspicious activity. Newer ML algorithms, such as XGBoost, LightGBM, and deep bolstering learning models, are applied for predictive analytics purposes to predict any potential vulnerability by going through historical incidents so that the firewall can adapt to address security risks even before a need for intervention arises. DL models, such as recurrent neural networks (RNNs), are applied to implement the capability of self-learning wherein the firewall could modify its responses according to evolving cyber threats. One example case is the predictive model of Singapore, where historical attack data across public infrastructure can be analyzed to help the city in finding its vulnerabilities earlier and reinforcement its digital defenses through accurate proactive anomaly threat detection.

During Phase 4, system integration and deployment, the AI firewall integrates into any other systems running in the digital city. This includes IoT networks, cloud services, and database systems that the departments of the smart city will use. Secure communication channels are established across the dissimilar systems, ensuring data are exchanged in real-time. The cohesive firewall is consequently released into a trial environment, preferably in a small city municipal service, where it will be subjected to assessments of its efficacy in stopping actual hacking attempts and stress tests pretending cyberattacks. Real-time alerts are established to alert the personnel once inconsistencies are detected, and automated retorts, such as isolating conceded devices, are established as part of protocols designed to contain the threat. This phase is effectively proven by the case study of London’s AI firewall pilot, where the firewall was integrated into the digital city traffic management system. Indeed, during initial deployment tests, the firewall identified unusual data patterns in real-time; it has the capacity to enhance urban cybersecurity.

Phase 5: In phase 5, monitoring, optimization, and constant learning, the intelligence firewall constantly monitors its performance metrics, such as the precision of threat detection and response time, but at the same time, it finds an incident response team to monitor major security events. As a result, computational learning models are regularly updated with new data to ensure the firewall remains strong in countering new terrorizations from the ever-changing cyber realm. Additional features of energy optimization are added as intelligence firewalls, automated detection of threats and resource management, which lowers the energy requirement to power monitoring systems. This includes the case study of New York City: In the smart city utility sector, intelligence firewalls were implemented to continue monitoring and automate model updates for data center energy efficacy optimization. This phase is expected to result in better in-time exposure of threats that will enrich cybersecurity, increased efficiency in operations through reduced intervention in the monitoring process of cybersecurity, and sustainability since optimized energy use also meets ecological goals without negotiating security.

The implementation of AI-driven firewalls in smart cities involves a multi-phase method that ensures not only enhanced cybersecurity but also the seamless operation of key services during cyber threats. The initial phase, Digital City Infrastructure Analysis and Security Planning, establishes the groundwork by assessing cyber risks and identifying vulnerabilities in key sectors such as traffic management systems, health care, and other utilities system. This phase ensures that AI-driven firewalls are designed to protect the most sensitive aspects of the urban landscape. A case study of Barcelona’s Smart City Initiative highlights this methodology, where a comprehensive cyber risk assessment was conducted to design and deploy effective AI-driven firewall solutions to protect essential services. Following this, the second phase, Data Acquisition and Preparation, focuses on collecting and deleting data from IoT devices across the smart city landscape. This data is preprocessed and organized into training datasets for the AI-driven models to ensure the precision and reliability of subsequent analyses.

The third phase, Advanced AI-driven Model Development and Adaptive Threat Prediction, involves developing ML methods that can identify threat anomalies and predict potential cyber threats. For instance, Singapore’s predictive analytics methods help detect vulnerabilities by analyzing historical data, allowing proactive prompt adjustments to the digital city’s security infrastructure. The fourth phase, System Integration and Secure AI Deployment, ensures that the AI-driven firewall seamlessly integrates with the city’s IoT networks and cloud systems. London’s deployment of an AI-driven firewall in its traffic management system, which detected unusual data patterns in real-time systems, enhancing urban cybersecurity. The fifth phase, Adaptive Monitoring and Sustainable Optimization, ensures continuous learning and adaptation by the AI-driven firewall. With features such as energy optimization and automated cyber threat detection, the AI firewall system maintains its efficiency while reducing operating costs. New York City’s implementation of an AI-driven firewall system for energy optimization in the utility sector is an example of how this phase also contributes to sustainability. These interconnected phases ensure the cybersecurity and operational continuity of the smart cities’ landscape, enabling them to be resilient to evolving cyber threats while maintaining their essential services.

Figure 2 shows a phased method for setting up an intelligence firewall system that can be used in a digital city. The development process begins with the city infrastructure analysis and security planning. Subsequently, data are collected and prepared to form the basis of a systematic process to identify the areas that are flaws for the smart city. The intelligence methods are designed to identify cyber risks and threats early and are integrated into systems within the digital city, thereby enabling the use of AI-driven technologies safely. Thus, methods that maintain security are established within the digital urban landscape through a constant monitoring and optimization system. This whole approach leads to the advent of a cyber secure and resistant type of sustainable digital framework [12, 13].

Figure 2:

Implementing an AI firewall in smart cities. AI, artificial intelligence.

This smart city firewall system algorithm, through AI, has five primary phases to secure the smart city milieu. The first phase in this system is Digital city setup analysis, which involves setting cybersecurity goals, evaluating cyber threats, and prioritizing shields for IoT systems, public essential services, and citizen data privacy. Data acquisition and preparation collect and progressions IoT and network data appropriately, with quality checks, into ready-to-use datasets for training. The next phase is the development of the intelligence method, which primarily involves analytical methods such as machine learning methods, such as K-means and RNNs, to track irregularities and impede threats using continuous learning techniques for adaptability. System integration and utilization involve integrating the firewall with urban systems, with mechanisms for testing cyber detection precision with automated reactions within the secure environment. To conclude, monitoring and optimization will guarantee that the security is always alive by measuring metrics so that models can have new data and be based on energy-efficient operations. Future enhancements will consider edge computing to provide better prompt response, promote improvement on privacy-preserving practices, and auto-incident retrieval that would ensure the firewall system scales well with the digital city’s infrastructure.

Figure 3 integrates the adoption trend of AI-driven firewalls through various stages in the lifecycle of a smart city system, which also marks KPIs in terms of extortion captured, prompt response time reduction, energy efficiency, and citizen reliance. In other words, each phase represents incremental novelty through the amalgamation process, incorporating the intelligence-driven firewall. For instance, with the evolution of smart city’s AI-driven firewall, an emergent trend witnessed is cyber threat detection and mitigation ability along with response time for cybersecurity breaches. Simultaneously, energy efficiency is improving due to the ongoing evolution of the AI firewall system into a much more optimized solution with fewer resource requirements, yet still ensuring a proficient level of security system. From Figure 3, it is inferred that there is a pattern for upward movement in the level of citizen reliance, which corresponds to a progressive increase in confidence with digital technological infrastructures in smart cities, which assure safety and confidentiality. This growth path describes the potential for AI-driven firewalls to allow smart cities to slowly, with time grow into safe, more responsive, and sustainable surroundings [14, 15].

Figure 3:

Performance metrics of AI firewall in smart cities. AI, artificial intelligence.

The rapidly evolving landscape of AI firewall system technology has introduced groundbreaking trends, particularly federated learning, and edge-driven AI approach, which promise to significantly enhance cybersecurity in smart cities landscape. Federated learning is a distributed ML method where AI-driven models are trained on data stored locally across different IoT devices rather than on a centralized server. This systematic approach has considerable potential for smart cities, where data are continuously generated by billions of IoT devices. By enabling the AI-driven firewall to learn from data across different urban systems while maintaining data privacy and security, federated learning reduces the risks associated with centralized security breaches. This method also improves the scalability of security systems, as AI-driven methods can be trained locally on edge AI-driven devices, ensuring faster learning and better adaptation to emerging cyber threats without relying on data transmission to central servers. With this decentralized strategy, federated learning can revolutionize how smart cities secure interlinked IoT devices while promoting data sovereignty and mitigating inherent cyber threats in cloud-based systems.

Edge AI models, however, involve processing data closer to its source, such as on IoT devices or on-premises servers, rather than sending it to distant cloud-based servers. This emergent trend enhances the real-time systems decision-making capabilities of AI firewall systems, enabling them to be more effective in dynamic smart city landscapes. With edge-driven AI models, smart city’s security systems can respond to cyber threats instantly, without the latency introduced by sending data to the cloud for analysis. In IoT environments where immediate or prompt action is crucial, such as autonomous vehicles, smart traffic light systems or public safety systems, edge AI allows for quicker cyber threat detection and mitigation, thus preventing potential cyberattacks before they occur. Moreover, edge AI provides a more effective utilization of network resources by reducing the need for constant data transmission, reducing bandwidth usage, and preserving data-driven energy. These advantages make edge AI a vital tool for ensuring the continuous operation of smart city landscapes, where delays in cybersecurity responses could have a prominent impact.

Both federated learning and edge-driven AI methods are transformative forces in the development of AI firewall systems for smart cities, paving the way for more resilient and versatile cybersecurity systems. These technologies empower AI-driven firewalls to learn and adapt faster and function with greater efficacy, all while maintaining data privacy and security. Integrating these emerging digital trends into smart city infrastructures promises to bolster the resilience of urban networks, protecting them from numerous cyber threats, ranging from cyberattacks on critical infrastructure to privacy breaches. However, the successful implementation of federated learning and edge-driven AI in AI firewall systems require overcoming obstacles such as data standardization, interoperability among diverse IoT devices, and the need for resilient protection frameworks in decentralized networks. As these advanced technologies continue to evolve, their combination with AI-driven firewall system will play a pivotal role in shaping the future of cybersecurity in smart cities, guaranteeing their safety, sustainability, and suitability to address the challenges of an increasingly digital world.

AI firewall systems are essential for ensuring the security and stability of smart cities, requiring a combination of innovative ML models and adaptive learning models. ML models such as supervised learning, support vector machines (SVMs) and Random Forests, are used to detect known cyber threats by identifying traffic patterns based on categorized data. In contrast, unsupervised ML methods such as K-means clustering and Isolation Forests are used to identify threat anomalies in traffic, which could indicate concealed threats. Semi-supervised ML techniques, which utilize both categorized and unlabeled data, are effective in real-world system scenarios with limited labeled data, while DL methods such as convolutional neural networks (CNNs) and RNNs tackle more complex and evolving cyber threats, such as advanced persistent threats (APTs) or zero-day vulnerabilities. Adaptive ML methods, such as reinforcement learning (RL), enable AI-driven firewall systems to independently adjust their response to emerging cyber threats, using algorithms such as Q-learning and deep Q-networks (DQNs) to optimize cryptographic security measures in real-time applications. Moreover, federated ML allows for decentralized learning across multiple smart cities enabled IoT devices, preserving data-driven privacy while still benefiting from shared knowledge, while transfer learning accelerates cyber threat detection by transferring knowledge from one landscape to another. For network traffic analysis, pattern recognition methods such as Naive Bayes, K-nearest neighbors (KNNs), and decision trees classify traffic based on predefined access models. In contrast, flow-based analysis methods, including NetFlow and sFlow, focus on monitoring network traffic to detect cyber anomalies such as DDoS attacks. Feature engineering techniques such as principal component analysis (PCA) system and recursive feature elimination (RFE) improve data processing efficacy, ensuring that only the most relevant features are considered, while data normalization safeguards that the data are consistent and suitable for model training. Given the continuous growth and complexity of the smart cities’ landscape, real-time system processing and scalability are essential, with edge-driven AI computing facilitating low-latency data processing closer to IoT devices, and distributed processing systems such as Apache Kafka and Apache Spark ensuring efficient handling of large-scale network traffic. By integrating these diverse ML models and adaptive models, AI firewall systems can dynamically learn, detect, and mitigate cyber threats, ensuring the robustness and security of critical smart city landscapes while supporting the scalability and sustainability of urban landscapes.

AI firewall systems are essential for enhancing cybersecurity in smart cities, which increasingly require interlinked IoT devices to manage essential infrastructure such as transportation system, energy, health care system, and public reliance. As smart cities become digitally transformable, traditional classical firewalls struggle to meet the growing complexity and sophistication of cyber threats. Unlike traditional classical firewalls, AI-driven firewalls utilize ML and DL algorithms to analyze network traffic in real-time system, detecting cyber anomalies, predicting potential cyber threats, and responding independently. This adaptability is essential in dynamic smart city landscapes, where rapid proliferation of new IoT devices and continuous evolution of cyberattack methods require a system capable of autonomously adapting without frequent manual updates. Beyond securing networks, AI-driven firewall systems also contribute to the sustainability and resilience of smart cities by safeguarding critical security systems, such as energy grids and waste management, thereby preventing disruptions that could affect essential services and citizen safety. They also enhance personal data privacy, ensuring that sensitive citizen information is protected from malignant individuals, thus fostering public reliance, and encouraging participation in digital city initiatives. This robust security and privacy framework is essential for citizens to confidently engage in the advanced technologies and services that are shaping their urban landscapes. However, the implementation of AI firewall systems in smart cities comes with its own set of obstacles, such as high initial case studies, ongoing maintenance requirements, and the need for specialized personnel, which can strain municipal budgets, particularly in smaller urban cities with limited resources. Moreover, AI-driven firewalls require access to enormous amounts of data for learning and adapting, resulting in concerns about citizen data privacy and potential misuse. Scalability and interoperability are also challenges, as the increasing complexity of urban systems requires AI firewall systems that can integrate seamlessly with various IoT devices and applications across multiple urban realms. Current research in AI firewalls is focused on overcoming these obstacles and enhancing their efficacy. Promising developments, such as federated ML, which enables decentralized data training without compromising data privacy, and edge-driven with AI firewall systems, which process data closer to the source to reduce latency and improve real-time system response times in applications such as traffic management and autonomous vehicles, are showing immense potential. Additionally, XAI is gaining momentum, offering transparency in AI-driven firewall decision-making, thus improving accountability, and ensuring compliance with data protection regulations. In conclusion, AI firewall systems are a systematic transformative solution for improving smart city’s security, efficiency, and sustainability. While obstacles remain in their deployment and integration, ongoing research and technological evolutions offer promising pathways for developing adaptive, scalable, and data-driven privacy-conscious cybersecurity solutions that can keep pace with the ever-evolving landscape of urban cyber threats.

To evaluate the effectiveness of AI-driven firewalls in improving the security and resilience of smart cities, it is essential to define measurable performance metrics that reflect real-world system scenarios. These performance metrics can be utilized to assess the AI-driven firewall’s effectiveness in maintaining security across the complex infrastructure of the smart cities landscape. The key performance indicators that provide cyber insights into AI-driven firewall performance are as follows:

Threat detection accuracy: The ability of the AI-driven firewall to accurately identify legitimate cyber threats while minimizing false positives (harmless data misclassified as threats) and false negatives (actual threats overlooked). Key metrics include precision (correctly identified threats), recall (identification of all actual threats), and the F1-score (a balance between precision and recall). High detection accuracy provides minimal disruption to critical urban services.

As smart city infrastructure evolves, the firewall system must scale to accommodate more devices and data. Throughput (data processed per unit time) and scalability index (performance with increasing devices and traffic) are key metrics to assess the firewall systems’ ability to adapt to growing smart urban environments.

The impact of AI-driven firewalls on the overall performance of city systems, particularly in high-demand scenarios, should not be underestimated. Latency impact and resource utilization (CPU, memory, and storage) are key indicators for assessing the firewall systems’ operational efficiency.

System reliability and availability: The AI-driven firewall’s ability to operate constantly without failure is essential. Metrics such as uptime percentage (how often the firewall is fully operational) and mean time between failures (MTBFs) are used to assess its consistency.

These performance metrics provide a comprehensive framework to assess AI-driven firewalls in smart cities, ensuring that they are not only protecting against cyber threats but also ensuring the long-term sustainability and effectiveness of urban infrastructure.

Integrating case studies into the paper can be beneficial in demonstrating the practical application of the proposed AI-driven firewall approach in enhancing the resilience and security of smart cities. Briefly, case studies will provide an overview of the implementation and efficiency of AI firewall systems in various smart city areas. For instance, Case Study 1 could focus on implementing AI-driven firewalls in the smart grid system. Here, the goal is to show how the AI firewall system can protect critical energy landscapes such as smart grids, in a smart city milieu. Challenges such as high data volume, real-time system monitoring requirements, and DDoS attacks on grid systems will be addressed by explaining how AI-driven firewalls utilize ML methods to detect abnormal traffic patterns and prevent cyberattacks. The results would provide measurable metrics, such as cyber threat detection precision and reduced downtime, highlighting the AI-driven firewalls’ success in mitigating real-world cyber threats. Case Study 2 examined the use of AI firewalls to secure health care networks. The obstacles would focus on sensitive data protection, compliance with regulations (e.g. HIPAA), and safeguarding interconnected medical devices. The AI firewall system would be used to continuously monitor network traffic, detect unauthorized access, and prevent cyberattacks, ensuring security and privacy for patient data. The results will highlight the improved detection rate of cyberattacks and a decrease in data breaches, leading to operational continuity in health care. Case Study 3 could address AI firewall systems in smart transportation systems. The obstacles, such as managing multiple IoT devices in transportation networks and cyber threats, such as GPS spoofing and vehicle control hacking, would be explored. The AI-driven firewall would apply DL algorithms to monitor traffic data and IoT interactions, detect cyber anomalies and prevent threats in real-time systems. Results would demonstrate reduced vulnerability to cyber threats, improved operational efficiency, and reduced delays in transportation services. Case Study 4 would focus on safeguarding public safety and surveillance systems in urban landscapes. Here, challenges such as large-scale video surveillance systems, citizen data privacy, and vulnerability of IoT-connected cameras and sensors would be discussed. AI-driven firewalls would safeguard surveillance systems by monitoring data flow and identifying potential cyber threats. Results would focus on improved security systems, quick cyber threat identification, and prevention of system downtimes in critical security networks. Overall, the research case studies demonstrate the successful implementation of the AI firewall system approach across various smart city systems, highlighting its feasibility and scalability. They would connect back to the research methodology and performance evaluation metrics, validating the effectiveness of the security system design, scalability, and adaptability to evolving cyber threats in dynamic smart city landscapes, proving that AI-driven firewalls are not just theoretical but are applicable and successful in real-world system scenarios.

Customized AI-driven firewalls for various urban settings must be based on the unique characteristics of each city’s landscape, governance model, and technological ecosystem. The firewall system design and functionality are influenced by factors such as the scale of IoT networks, volume and sensitivity of data, and specific cyber threat landscapes. Collaboration with municipal authorities, private enterprises, and cybersecurity experts ensures consistency with local requirements while addressing global standards. Standard benchmarks and guidelines, such as the NIST Cybersecurity Framework, ISO/IEC 27001 for information system security, and IEEE standards for IoT security, provide foundational approaches for effective implementation. These Customized AI Firewall frameworks emphasize real-time cyber threat detection, adaptive learning, scalability, and data privacy, enabling AI firewalls to protect critical infrastructure while complying with legal and regulatory requirements. Integrating modular and interoperable designs ensures that the firewalls remain viable and adaptable to emerging city challenges.

The enactment of AI-driven firewalls in smart cities requires significant upfront investment and operational costs, but their long-term benefits make them a strategic choice to enhance urban robustness and sustainability. The initial costs include buying advanced hardware, developing, and training AI-driven models, acquiring software licenses, integrating firewall systems into existing infrastructure, and establishing edge computing nodes for real-time system threat processing. Ongoing expenses include updating the system, retraining AI-driven models, and maintaining federated learning systems. Despite these costs, the advantages of the AI-powered firewall system surpass the financial burden when considering its role in protecting essential urban services. These firewall systems leverage advanced cyber detection techniques such as supervised ML, anomaly detection, and adaptive DL to minimize the frequency and impact of cyberattacks, preventing service disruptions in critical sectors such as transportation systems, energy, and health care system. The financial savings from avoiding cyberattacks, such as a data breach in health care system, which can cost millions of dollars in penalties and recovery costs, make these systems indispensable. Furthermore, AI-driven firewalls improve operational efficiency by automating cyber threat detection and mitigation, reducing the reliance on human intervention and labor costs, while ensuring continuous monitoring and prompt response. These systems also enhance compliance with cybersecurity standards such as GDPR, ISO 27001, and NIST, preventing costly penalties for non-compliance. Indirectly, AI-driven firewalls enhance public reliance by ensuring secure data handling and continuous urban services, leading to greater adoption of digital city technologies and overall improvements in the quality of life of citizens. Moreover, AI firewall systems provide scalability, adapting to evolving cyber threats and expanding infrastructure, ensuring long-term efficiency even as cities grow and integrate more IoT devices. When evaluating the ROI, factors such as reduced incident prompt response costs, savings from minimized downtime, compliance achievements, and increased citizen engagement highlight the cost-effectiveness of the AI-powered firewall system in sustaining smart city operations. Although the initial financial expenditure is considerable, the long-term benefits of cybersecurity, operational continuity, and resilience make AI-driven firewall system an essential investment for the future of the smart cities’ landscape.

Table 4 highlights key performance metrics of AI-driven firewalls in the smart cities’ landscape, focusing on their efficiency. AI-driven firewalls provide excellent cyber detection precision, ranging from 97.5% to 98%, ensuring the rapid identification and mitigation of cyber threats. Their prompt response time is exceptionally fast, typically under 1 s, enabling immediate retaliation against system security incidents. Moreover, AI-driven firewalls contribute to energy efficiency, reducing power consumption by up to 30% through an optimized resource management system. The proactive cyber threat detection capabilities of these systems enable the early detection of vulnerabilities, enhancing overall cybersecurity posture by detecting cyber threats up to two weeks in advance. Furthermore, AI firewall systems demonstrate cost efficiency, with a 25% reduction in cybersecurity-related costs in the first year due to lower service downtime and reduced manual intervention. It is noteworthy that AI-driven firewall system can secure networks involving more than 500,000 devices without requiring performance degradation, ensuring their adaptability to the growing demands of smart city landscapes. These performance metrics highlight the crucial role AI firewall systems play in enhancing smart cities’ security, sustainability, and operational efficiency.