Using ethical artificial intelligence (EAI) to achieve sustainable development, Iraq as a case study

Abstract

Introduction:

Iraq faces persistent challenges in achieving sustainable development due to decades of conflict, political instability, and infrastructural degradation. These challenges are particularly evident in critical sectors such as energy, water, healthcare, education, and governance, which significantly influence human well-being, social equity, and quality of life. This study proposes an AI-driven, ethically guided, and human-centric sustainability framework to support resilient urban transformation in Iraq.

Methods:

The proposed framework integrates Ethical Artificial Intelligence (EAI), machine learning techniques, and a computational decision-support system (DSS). A hybrid modeling approach combining Multi-Criteria Decision Analysis (MCDA) and AI is developed to evaluate sustainability performance across interconnected sectors, including clean energy, water security, smart transportation, environmental protection, e-governance, and human development. The system incorporates real-time data analytics and a customized software prototype adapted to Iraq’s socio-economic and environmental context. Ethical principles such as transparency, fairness, accountability, privacy protection, and bias mitigation are embedded throughout the model design and implementation.

Results:

The framework enables dynamic and real-time sustainability assessment across multiple urban sectors. When applied to the Baghdad case study, it demonstrates improved performance in energy distribution efficiency, water resource management, healthcare service delivery, and governance transparency. The results indicate enhanced decision-support capability and optimized resource allocation, while explicitly prioritizing human development indicators within the evaluation and optimization process.

Discussion:

The findings highlight the potential of Ethical Artificial Intelligence as a transformative enabler of the United Nations Sustainable Development Goals (SDGs), particularly SDGs 3, 4, 6, 7, 9, 10, 11, and 16 in post-conflict contexts. The proposed framework provides a scalable and transferable model for sustainable urban transformation. It further demonstrates that embedding Ethical AI as a governing layer is essential for ensuring transparency, equity, accountability, and long-term resilience in smart city systems.

Graphical Abstract

1 Introduction

Iraq, a country with a rich historical, cultural, and civilizational heritage, has endured decades of conflict, political instability, and economic challenges, which have profoundly affected its infrastructure, social services, and environmental sustainability. Despite ongoing reconstruction and development efforts, the nation continues to face critical shortages in electricity and water, inadequate healthcare and education systems, environmental degradation, inefficient transportation networks, and governance challenges. These issues not only hinder economic growth but also compromise human well-being, social resilience, and progress toward the United Nations Sustainable Development Goals (SDGs).

Traditional development approaches in Iraq have largely relied on fragmented datasets, reactive policy measures, and sector-specific interventions, which remain insufficient to address the complex interdependencies among energy, water, environment, human development, governance, and urban planning. Consequently, decision-makers often struggle to prioritize interventions, optimize resource allocation, and anticipate emerging socio-economic and environmental needs. This highlights the absence of an integrated, predictive, and data-driven decision-support system capable of synthesizing multi-sectoral information and enabling proactive planning.

In this context, artificial intelligence (AI) emerges as a transformative tool capable of integrating heterogeneous datasets, forecasting sectoral demands, and optimizing resource distribution to support evidence-based and anticipatory decision-making. Accordingly, this study develops an AI-driven, human-centric framework designed to accelerate sustainable urban development in Iraq by improving decision-making across critical sectors, including electricity, water, healthcare, education, environment, transportation, governance, and security. The framework is guided by a central research inquiry: how can AI be effectively integrated within an ethically grounded and human-centric system to enhance sustainable development outcomes in post-conflict urban environments such as Iraq?

To address this question, the study investigates interconnected sectoral challenges and explores how AI-based predictive models and multi-criteria decision analysis (MCDA) can support optimized planning and resource allocation. The proposed framework specifically aims to improve electricity and water management through demand forecasting and system optimization, enhance healthcare and education services through equitable resource distribution, strengthen environmental sustainability through pollution monitoring and mitigation, and improve governance through data-driven transparency and accountability mechanisms. In addition, it addresses transportation inefficiencies through mobility modeling and congestion reduction strategies, while also incorporating security and stability considerations through risk prediction and urban resilience analysis.

A defining feature of this research is its human-centric approach, which prioritizes human development outcomes—including health, education, access to essential services, and social equity—within all analytical and optimization processes. This is operationalized through weighted multi-sectoral indicators embedded in the mathematical model, ensuring that AI-driven recommendations are evaluated not only based on infrastructural efficiency but also on their direct impact on human well-being and societal development.

The significance of this study lies in its ability to bridge fragmented urban systems through an integrated analytical architecture that combines artificial intelligence, mathematical modeling, and MCDA into a unified decision-support framework. This enables scenario-based simulation, predictive analytics, and dynamic optimization across interdependent urban sectors, thereby improving strategic planning, institutional efficiency, and policy responsiveness in complex post-conflict environments.

The novelty of the research is further emphasized by its context-specific application to Baghdad, a high-density, post-conflict metropolitan system that represents the broader challenges faced by Iraqi cities. By utilizing real-world multi-source datasets and aligning the framework with the United Nations Sustainable Development Goals (SDGs), the study ensures both empirical relevance and global applicability. Although validated in Baghdad, the proposed framework is scalable and transferable to other cities experiencing similar socio-economic and infrastructural constraints, offering a practical roadmap for resilient, inclusive, and sustainable urban recovery.

1.1 Ethical artificial intelligence (EAI)

Ethical Artificial Intelligence (EAI) plays a central role in the proposed framework by ensuring that all AI-driven decisions fully align with the Sustainable Development Goals (SDGs) and with internationally recognized ethical standards. The study integrates principles of fairness, transparency, accountability, data privacy, inclusiveness, and non-discrimination into every analytical stage—from data preprocessing to model interpretation. This ethical layer guarantees that AI predictions do not reinforce existing inequalities, do not bias decisions against vulnerable populations, and remain consistent with SDG priorities such as Quality Education (SDG 4), Clean Water and Sanitation (SDG 6), Affordable and Clean Energy (SDG 7), Sustainable Cities and Communities (SDG 11), Climate Action (SDG 13), and Peace, Justice, and Strong Institutions (SDG 16).

By embedding EAI principles into the multi-sector model, the framework ensures that technological advancement directly contributes to human-centered, equitable, and sustainable development outcomes in Iraq. The integration of EAI thus strengthens the credibility, transparency, and social acceptability of AI-driven recommendations, ensuring that the proposed framework remains aligned with national development needs while safeguarding ethical integrity.

2 Literature review

Artificial Intelligence (AI) has emerged as a transformative enabler of the United Nations Sustainable Development Goals (SDGs), offering innovative pathways to address social, economic, and environmental challenges across multiple sectors. Recent studies highlight the versatility of AI applications, ranging from optimizing energy consumption and advancing healthcare services to improving educational outcomes, promoting environmental sustainability, and supporting urban development (Waheeb et al., 2020; Waheeb and Andersen, 2022; Waheeb et al., 2023; Marjan, 2023; Alsaigh et al., 2022; Alalaq, 2025; Efremova et al., 2019; Salah, 2025; Goh, 2021). Internationally, AI has been applied in Singapore’s Smart Nation initiatives to optimize energy grids and urban services (Smart Nation and Digital Government Office, 2020), and in Dubai’s Smart City programs to integrate AI for water management and traffic optimization (Al Nuaimi et al., 2015). Predictive analytics, for instance, have been applied to forecast energy demand, reduce wastage, and integrate renewable energy sources, while AI-driven healthcare tools enhance diagnostics, personalize treatment protocols, and manage healthcare resources. In education, adaptive AI technologies facilitate tailored learning experiences, thereby improving knowledge retention, reducing dropout rates, and strengthening human capital formation. Globally, adaptive learning AI platforms like Knewton and Smart Sparrow have demonstrated improved student outcomes and retention rates across multiple educational systems (Luckin et al., 2016).

In the domains of energy and water management, AI plays a pivotal role in improving efficiency, reliability, and sustainability. Machine learning models can predict consumption trends, optimize grid operations, and integrate renewable sources to mitigate shortages. AI also supports leak detection, demand forecasting, distribution optimization, and wastewater treatment, contributing to both resource conservation and cost reduction (Waheeb et al., 2020; Waheeb and Andersen, 2022; Waheeb et al., 2025a,b). International case studies include the European Smart Grid projects that employ AI to balance renewable energy supply and demand across multiple urban centers (European Commission, 2021), and Japan’s AI-enabled water networks that detect leaks and optimize pumping schedules in real-time (Mori et al., 2020). These capacities are particularly critical in fragile contexts like Iraq, where long-standing conflict and underinvestment have weakened infrastructure, making AI adoption essential for efficient public service delivery.

Globally, the integration of artificial intelligence into healthcare systems has demonstrated improvements in diagnostic accuracy, workflow efficiency, and institutional performance, particularly in stable and high-income settings (Topol, 2019a,b; Rajpurkar et al., 2017). In contrast, in post-conflict and fragile contexts such as Iraq, healthcare systems are constrained by infrastructural damage, workforce shortages, and governance challenges, and empirical evidence supporting AI-driven improvements in patient outcomes remains limited. Current applications are mainly used as decision-support tools for prioritization and resource allocation rather than as fully outcome-validated systems (Waheeb et al., 2023; Marjan, 2023; Al-zubidi et al., 2024; Guidance, W. H. O, 2021; Leslie et al., 2021). Globally, AI-driven systems such as IBM Watson Health and DeepMind’s medical imaging tools have been deployed to predict patient deterioration, optimize treatment, and improve hospital workflow efficiency (Topol, 2019a,b; Rajpurkar et al., 2017). Similarly, adaptive learning strategies enhance educational outcomes by tailoring content to student capabilities (Marjan, 2023; Waheeb et al., 2023).

AI also contributes substantially to environmental sustainability. It enables real-time pollution monitoring, improved waste management, and biodiversity protection. Additionally, AI supports smart city development by enabling energy-efficient buildings, intelligent transportation systems, and resilient urban infrastructure. In Iraq, AI-driven urban solutions have been applied to transform Baghdad into a Smart and Sustainable City, optimizing infrastructure, enhancing monitoring systems, and improving public services through ethical and context-sensitive frameworks (Waheeb et al., 2025a,b; Mohammed and Altaie, 2025). Internationally, cities like Copenhagen and Amsterdam employ AI for energy-efficient building management, pollution prediction, and sustainable waste collection, serving as global benchmarks for urban sustainability (Angelidou, 2015; Jabareen, 2006).

Transportation systems have benefited from AI applications designed to optimize traffic flow, reduce congestion, and support sustainable mobility. In governance, AI strengthens decision-making through improved transparency, accountability, and evidence-based policymaking (Waheeb and Andersen, 2022; Alalaq, 2025). Examples include AI-driven traffic management in Tokyo and AI-assisted policy modeling in smart European cities, which have reduced congestion and improved urban governance efficiency (Batty et al., 2012; Nam and Pardo, 2011). Despite challenges such as weak infrastructure, political instability, and limited technical capacity, Iraq continues to demonstrate growing academic and institutional interest in AI applications for sustainability, post-disaster reconstruction, and climate resilience (Waheeb et al., 2020; Altaie and Dishar, 2024; Altaie et al., 2023).

A number of sector-specific studies provide additional depth. Altaie and Borhan (2019) emphasized the need for AI-integrated construction management systems to reduce time and cost deviations and enhance project stability, especially in fragile environments like Iraq. Mohammed and Altaie (2025) highlighted the necessity of a unified smart city framework that integrates themes of inclusivity, standardization, and social justice. Mansour and Salman (2025) demonstrated the effectiveness of localized sustainability practices—such as renewable-energy-powered recycling systems—in reducing institutional costs and carbon emissions. Yaseen and Rasoqi (2025) argued that sustainable urban contexts foster community-oriented and livable environments through balanced physical, social, and environmental design criteria.

Cybersecurity applications were advanced by Al-zubidi et al. (2024), who developed a hybrid CNN-LSTM-XGBoost model for detecting DoS and DDoS attacks with exceptional accuracy (98.3%–99.3%), reflecting the potential of AI in safeguarding digital infrastructure. Internationally, AI-based cybersecurity tools, including Palo Alto Networks Cortex XDR and Microsoft Azure Sentinel, are deployed to predict and mitigate cyber threats in urban infrastructures (Buczak and Guven, 2016).

Waheeb et al. (2025a,b) provided two notable contributions: a comprehensive, ethical, AI-based framework for transforming Baghdad into a smart and sustainable city, and an expert-system-driven methodology for addressing cost overruns and delays in post-disaster reconstruction. Their expert system incorporated 80 causes of delay categorized into major and secondary factors, revealing that contractors, project owners, external factors, consulting teams, and administrative systems were the primary contributors to delays.

Altaie and Dishar (2024) examined reconstruction project management in Mosul, Anbar, and Tikrit. Using the Analytical Hierarchy Process (AHP), they prioritized contributing factors to challenges in scope, cost, time, and quality. Scope change emerged as the most significant challenge (40.8%), followed by cost (27.6%), quality (18.11%), and time (13.5%). Their study developed five mathematical equations to guide decision-makers in assessing and managing these factors, recommending realistic administrative and methodological strategies to enhance project outcomes.

Ethical AI frameworks have recently improved urban governance efficiency (Das, 2024). AI-supported decision systems enhance sustainability planning and smart city management (Batarseh et al., 2023; Homaei et al., 2024). Furthermore, advanced predictive models in AI have demonstrated strong performance in healthcare and data-driven decision-making (Rajpurkar et al., 2022; Buaka and Moid, 2024).

Finally, Altaie et al. (2023) examined post-disaster reconstruction challenges by surveying 43 experienced engineers and identifying 49 delay-related factors. Using the Relative Importance Index (RII), they ranked four main criteria—scope, time, cost, and quality—and identified 13 critical factors that significantly affect reconstruction outcomes. Their findings underscore the need for proactive decision-making and local capacity-building to support sustainable rebuilding efforts, reduce displacement, and restore socio-economic stability.

Table 1 illustrative examples of Artificial Intelligence (AI) applications across major sustainable development sectors, outlining their key contributions, representative and international references, and their alignment with the United Nations Sustainable Development Goals (SDGs), with emphasis on sector-specific indicators, measurability, and real-world applicability. The table is illustrative rather than exhaustive and presents representative examples of how AI applications may align with selected SDGs, particularly in post-conflict and fragile contexts where context-specific adaptation is essential.

3 Methodology

This study adopts an integrated, ethical AI–driven methodological framework to support sustainable development decision-making in Iraq, with Baghdad serving as a representative urban case study. Unlike conventional sector-specific approaches, the proposed methodology is designed to capture interdependencies and feedback mechanisms among critical development sectors, including electricity, water, healthcare, education, transportation, environment, security, and governance.

As illustrated in Figure 1, the workflow combines sector-specific artificial intelligence models within a unified decision-support architecture, enabling data sharing, cross-sectoral interaction, and scenario-based optimization. Ethical AI principles—such as transparency, fairness, accountability, and human-centered decision-making—are embedded throughout the modeling process to ensure responsible and context-aware policy support.

Figure 1

The methodological process comprises four main stages: (i) multi-source data collection and preprocessing, (ii) sectoral AI modeling and validation, (iii) integrated simulation and optimization across sectors, and (iv) impact assessment and scenario analysis. Detailed information on the algorithms used, input variables, data volume, validation techniques, performance metrics, and error rates for each sector in Table 1, while the quantitative cross-sectoral impacts derived from the integrated framework are presented in Table 2.

Figure 1 presents the methodological workflow of the proposed AI-driven sustainable development framework for Baghdad, Iraq. The figure illustrates the sequential process of multi-sectoral data collection, data preprocessing, AI-based prediction and optimization (including MCDA), scenario simulation, and decision-support software outputs. This workflow enables evidence-based sustainability assessment and actionable policy recommendations to support post-conflict reconstruction and long-term urban resilience.

3.1 Clarification of claims

To avoid overstatement, the claims regarding “actionable insights,” “operational feasibility,” and a “scalable roadmap” have been refined. The contributions of this study are positioned as decision-support–oriented rather than implementation-prescriptive.

Specifically, the proposed AI-driven framework provides policy-relevant and evidence-based guidance through predictive analytics, scenario simulation, and multi-criteria evaluation. While the framework supports prioritization, comparison of policy options, and strategic planning, it does not constitute a fully detailed implementation or deployment roadmap. Instead, it offers a conceptually operational and analytically validated decision-support structure that can inform future, context-specific implementation planning by relevant institutions.

In addition, references to “global sustainability standards” are explicitly aligned with the United Nations Sustainable Development Goals (SDGs) and their associated indicators, which are used as the primary benchmark for evaluating sustainability dimensions within the framework.

3.2 Methodology framework

The integrated framework incorporates eight primary units critical for post-conflict urban recovery. Detailed modeling is applied to Electricity, Water Supply, Healthcare, and Education, while Transportation, Environmental Management, Governance, and Security/Safety are represented in a simplified form due to data and scope limitations. Ethical AI is used to coordinate cross-sectoral information, ensuring that human development priorities guide decision-making.

The integrated framework incorporates eight key sectors critical for urban recovery in Baghdad. Electricity, Water Supply, Healthcare, and Education are modeled in detail due to their direct influence on human development outcomes. Transportation/Traffic, Environmental Management, Governance/Smart City, and Security & Occupational Safety are represented in a simplified form due to data availability and analytical scope constraints. Ethical AI coordinates cross-sectoral information to support human-centric sustainable urban development decision-making (see Figure 2).

Figure 2

3.3 Justification of human development dimensions and ethical framing

The selection of human development factors in this study—namely healthcare, education, access to water and energy, security, and social equity—is conceptually grounded in established international development and ethical frameworks rather than project-specific choices. These dimensions align directly with the core components of the Human Development Index (HDI), which emphasizes health and education as fundamental pillars of human well-being (UNDP, 2020), as well as with key United Nations Sustainable Development Goals (SDGs), particularly SDGs 3 (Good Health and Well-being), 4 (Quality Education), 6 (Clean Water and Sanitation), 7 (Affordable and Clean Energy), 11 (Sustainable Cities and Communities), and 16 (Peace, Justice, and Strong Institutions) (United Nations, 2015).

In post-conflict and fragile settings, existing recovery and resilience literature consistently identifies these dimensions as foundational preconditions for stabilization and human security, rather than as downstream development outcomes. Access to healthcare, education, basic services, and security is widely recognized as essential for restoring social trust, institutional legitimacy, and minimum living standards before more advanced development objectives—such as livelihood diversification, housing quality, or gender empowerment—can be effectively pursued (World Bank, 2017; UNDP, 2018; UN-Habitat, 2020).

Accordingly, dimensions such as livelihoods, housing, and gender equality were not excluded due to a lack of importance, but were considered secondary or dependent variables within the scope of the current AI-based decision-support framework. Their outcomes are strongly mediated by improvements in the primary human development factors modeled in this study. This scope definition reflects both analytical focus and data availability constraints at the district level, which are common limitations in post-conflict urban environments (World Bank, 2017).

From an Ethical AI perspective, the term human-centric is explicitly aligned with recognized normative standards, including the EU Ethics Guidelines for Trustworthy AI (European Commission, 2019), UNESCO’s Recommendation on the Ethics of Artificial Intelligence (UNESCO, 2021), and the OECD AI Principles (OECD, 2019). Within this framework, human-centricity is operationalized through criteria that prioritize human well-being, social inclusion, non-discrimination, and the protection of fundamental human needs, ensuring that AI-driven decision support serves societal recovery objectives in post-conflict urban contexts.

Waheeb et al. (2025a,b) demonstrate that integrating artificial intelligence with urban governance enhances resilience, sustainability, and data-driven decision-making in fragile and post-conflict contexts.

3.4 Research design

This study adopts a mixed-methods research design, combining quantitative and qualitative approaches to develop a comprehensive AI-driven framework for sustainable development in Iraq. The methodology integrates data collection, mathematical modeling, computational programming, predictive analytics, and scenario simulation to construct a robust decision-support system (DSS) that guides human-centric development.

The design ensures alignment between technical solutions and social objectives, addressing eight critical sectors: electricity, water, healthcare, education, environment, transportation, governance, and security/safety. By integrating predictive analytics, optimization models, and expert-driven weighting of priorities, the framework provides actionable insights for policymakers and urban planners, taking into account resource constraints, population density, and post-conflict realities.

3.5 Human-centric operationalization

In this study, the human-centric approach is operationalized by explicitly defining and weighting human development factors, including healthcare, education, access to water and energy, security, and social equity. Weighted scores are assigned to each sectoral indicator—energy, water, environment, education, healthcare, transportation, governance, and security—within the mathematical optimization model. Scenarios are evaluated to maximize human development outcomes while balancing infrastructure efficiency and resource allocation. This framework ensures that AI-driven interventions prioritize human well-being across all sectors, rather than focusing solely on system performance or technical efficiency.

3.6 Data collection and case study

This study employs a multi-source data approach to analyze the application of Ethical AI for sustainable development in Iraq, with Baghdad as the primary case study. Real-world data were collected from Iraqi ministries, governmental agencies, and field surveys to provide a comprehensive overview of the country’s development landscape.

The datasets cover multiple districts of Baghdad and span the period 2015–2024. Data were obtained from the Iraqi Ministry of Electricity, Ministry of Water Resources, Ministry of Health, Ministry of Planning, municipal records, and official security reports, supplemented by field surveys. All datasets were cross-validated against official statistics and international reports where available.

Energy and Water: Hourly electricity consumption, solar generation, peak demand periods, water demand per district, and supply availability were collected from the Iraqi Green Climate Organisation/Ministry of Environment (2022), Baghdad Electricity Distribution Company (2022), Ministry of Water Resources, Baghdad Water Directorate (2023), and the U.S. Energy Information Administration (2025).

Transportation: Urban mobility patterns, traffic counts, metro ridership, and congestion levels were obtained from Baghdad Transport Authority (2022) and Transportation and Communication Statistics Directorate – Iraq (2023).

Healthcare and Education: Hospital capacities, medical staff numbers, patient loads, student enrollment, teacher counts, and student–teacher ratios were collected from the Ministry of Health – Baghdad Hospitals (2022), Arab NGO Network for Development (2023), Baghdad Directorate of Education (2022), and UNICEF (2020).

Environment: Environmental indicators including AQI, recycling percentages, and biodiversity records were obtained from the ADYAN Foundation (2024) and Al-Din et al (2024).

Security and Governance: Data on crime rates, police distribution, emergency response coverage, public service completion rates, and complaint resolution were sourced from IOM (2020), United Nations Development Programme (2024), International Organization for Migration (2020), Federal Commission of Integrity (Republic of Iraq) (2024), Baghdad e-Government Portal (2023), and RTI International (2013).

3.7 Primary case study – Baghdad

Baghdad, Iraq’s largest urban center, faces post-conflict reconstruction challenges, high population density, and significant infrastructural stress. Secondary benchmarks were drawn from other post-conflict and rapidly developing regions, including Hiroshima (Japan) and selected Gulf countries, to validate and adapt the AI-driven sustainable development framework.

3.8 Integration with AI smart model

The collected datasets across all sectors were fed into an Ethical AI Smart Model, allowing simulation, optimization, and decision support for sustainable urban development (Figure 3).

Figure 3

Figure 3 Ethical AI smart model for sustainable development in Iraq: an integrated framework covering eight core sectors—energy, water, transportation, health, governance, security, education, and environment. The model illustrates the central role of ethical AI in coordinating multi-sector data, enabling decision support, and promoting human-centric sustainable development aligned with SDGs.

3.9 Integration with AI smart model

To operationalize Ethical AI principles within the framework (Figure 3), all collected datasets across eight sectors—Energy, Water, Transportation, Healthcare, Governance, Security, Education, and Environment—undergo data quality checks, bias assessment, and protected-group analysis to ensure fairness and representation. The hybrid MCDA-AI model incorporates human-centric weighting of indicators and applies constraints to prevent disproportionate influence from any single sector or social group. Model outputs are accompanied by interpretability reports, transparency statements, and protected-group outcome analyses, enabling decision-makers to understand how recommendations are generated, while ensuring accountability, non-discrimination, and alignment with sustainable urban development goals. These steps demonstrate that Ethical AI is actively integrated across data governance, model computation, and DSS outputs, rather than being merely declarative.

Table 2 provides a structured overview of the data sources used in the Baghdad (Iraq) case study and links each sector with the specific variables collected and their corresponding in-text citations. The table clarifies the origin of each dataset—whether from governmental agencies, international organizations, or published academic work—and ensures full transparency and traceability of the empirical evidence supporting the analysis.

3.10 Innovative AI-mathematical model

3.10.1 Rationale for adopting the mathematical model

The decision to employ this integrated model was driven by three methodological considerations.

First, sustainable development in Baghdad spans multiple sectors—including energy, water, transportation, healthcare, education, environment, security, and governance—each using different scales and measurement units. A composite mathematical formulation was therefore required to harmonize these heterogeneous indicators into a single quantitative index.

Second, long-term planning in Baghdad requires predictive, rather than purely descriptive, analytical capability. Historical data alone cannot capture future electricity shortages, water stress, congestion patterns, security risks, or environmental decline. Incorporating machine learning functions enables the model to generate district-level forecasts, providing proactive insights for policymakers.

Third, Baghdad operates under structural constraints such as limited budgets, uneven population density, and unequal service distribution. Optimization techniques—both linear and nonlinear—were integrated to improve resource allocation, reduce shortages, and enhance the efficiency of sectoral interventions.

3.10.2 Justification for the functions used

The model employs normalized predictive functions for electricity and water demand, transforming raw indicators into a standardized 0–1 scale while embedding AI-based forecasts. This enables variables with different units (e.g., kilowatt-hours, cubic meters, hospital beds, crime counts) to be incorporated into a single composite metric. Combining MinMax normalization with AI outputs enhances both accuracy and interpretability.

The machine learning component was incorporated due to the complex and non-linear relationships among urban sustainability indicators. Regression models support demand forecasting, artificial neural networks capture deeper non-linear interactions, ensemble approaches such as Random Forest and Gradient Boosting enhance robustness, and scenario simulation allows policymakers to examine alternative futures and stress-test interventions.

3.10.3 Mathematical formulation

The overall sustainability score SSi for district i is computed as:

Where w1, w2,…, w8 are weights determined via expert consultation, and f(Ei),f(Wi) are normalized predictive functions from AI models. This formulation allows integration of quantitative forecasts and qualitative assessments, reflecting both technical constraints and social development priorities.

3.11 Indicator selection and weighting rationale

For each sector, multiple candidate indicators were initially identified based on international standards, including the UN Sustainable Development Goals (SDGs), World Bank urban development indicators, and ISO 37120. Indicators were selected according to three criteria: (i) relevance to post-conflict urban sustainability in Baghdad, (ii) data availability and reliability, and (iii) measurability at the district level. Sector-specific indicators were aggregated into composite indices (Hi, Edi, Envi, Ti, Gi, Si) using normalized scores.

Weights (w1–w8) were determined through expert consultation and Multi-Criteria Decision Analysis (MCDA), considering urgency, feasibility, and social impact.

Objective Function (Optimization):

Subject to sectoral constraints (electricity, water, security units, budget limits). Solvers: PuLP (linear/integer) and SciPy minimize (nonlinear).

3.11.1 Integration of AI models into the mathematical framework

The model follows a structured workflow. Data preprocessing includes KNN-based imputation for missing values and IQR-based outlier detection. Machine learning models—specifically ANN and Random Forest—generate district-level predictions for electricity and water demand. These predictions are normalized using MinMax scaling with uncertainty adjustment to ensure stability.

The final composite sustainability score synthesizes all sectoral indicators into a single decision-support metric. This enables district ranking, identification of critical weaknesses, targeted budget allocation, and prioritization of sectoral interventions. The integration of AI with MCDA and mathematical optimization strengthens the methodological rigor of the model and enhances its suitability for planning in resource-constrained urban environments.

3.12 Computational program (decision-support software)

A custom decision-support software platform was developed using Python to integrate multi-sectoral data and advanced analytics for sustainable development planning in Iraq. The platform incorporates data preprocessing modules to clean, normalize, and integrate multi-source datasets, ensuring reliability and consistency across sectors. Embedded AI engines, including neural networks, regression, and classification algorithms, provide predictive analytics for electricity, water, healthcare, education, environment, transportation, governance, and security. Optimization solvers, implemented through linear and non-linear programming modules using PuLP and SciPy, enable efficient allocation of resources to minimize shortages and enhance equity and resilience. The platform also features interactive visualization dashboards built with Dash/Plotly, which display real-time sustainability scores, scenario comparisons, sectoral gaps, and risk heatmaps for transportation and security. Policymakers can input real-time updates, run multiple simulations, compare alternative strategies with predicted outcomes, and generate exportable reports to support stakeholder engagement and strategic planning. By integrating AI-driven analytics with optimization and visualization tools, the software provides actionable insights that enhance evidence-based policy, improve service delivery, and strengthen Iraq’s post-conflict development resilience.

Ethical AI principles were operationalized as explicit constraints within the AI decision-making layer of the proposed framework. These constraints ensured transparency, fairness, accountability, and human-centered decision-making, thereby preventing purely efficiency-driven optimization outcomes.

3.13 Workflow and implementation steps

The study follows a structured workflow summarized in Figure 2. The analytical framework integrates multi-source data collection, advanced AI-driven predictive modeling, optimization techniques, scenario simulation, and software implementation into a coherent decision-support system. Each step contributes to improving sectoral efficiency, resource allocation, and strategic policymaking, enabling evidence-based decisions that address infrastructural, social, environmental, and security challenges, thereby promoting equitable and sustainable development across all sectors.

To ensure methodological clarity and transparency, Table 3 presents a comprehensive overview of the proposed framework, detailing its sequential phases, analytical techniques, functional objectives, and associated impacts on sustainable development. The framework combines multi-source data integration, advanced artificial intelligence (AI) modeling, optimization methods, and scenario-based simulations to enhance decision-making efficiency, particularly within post-conflict and fragile environments such as Iraq.

Table 3 comprehensive methodological framework outlining the sequential phases, analytical techniques, and expected outputs of the proposed AI-driven sustainable development model, spanning data acquisition, preprocessing, mathematical modeling, optimization, scenario simulation, system implementation, and policy formulation. Each phase is linked to its functional objectives and corresponding impact on sustainable development, with particular relevance to post-conflict and fragile environments. The framework emphasizes integration of multi-source data, advanced AI techniques, and decision-support tools to enhance accuracy, efficiency, and resilience in planning and governance.

This table is designed to enhance methodological transparency and reproducibility. Reported crime hotspot percentages (22%–27%) are based on United Nations Development Programme (2024) and International Organization for Migration (2020). The predicted improvement rate (+18%) is generated using neural network and ensemble learning approaches, incorporating key variables such as hotspot density, surveillance systems, and response time efficiency. This clarification directly addresses reviewer concerns regarding data sources and the computational basis of predictive outcomes.

3.13.1 Example of machine learning application

In the electricity sector, a regression model predicts daily electricity demand for each district based on historical consumption data, population density, and weather conditions:

The predicted demand is then used in the optimization module to allocate electricity efficiently across districts. Similarly, neural networks predict healthcare service gaps, and ensemble learning combines multiple sectoral models to improve overall predictive accuracy.

Figure 4 software architecture flowchart of the smart city decision-support system, depicting the end-to-end data pipeline from multi-source data acquisition and preprocessing to AI-driven modeling, optimization, and scenario simulation, culminating in interactive dashboards and decision-support outputs for policymakers.

Figure 4

3.14 Methodological clarification and analytical model

This study employs a mixed-methods analytical framework that combines quantitative statistical analysis with AI-based modeling to evaluate and enhance sustainable development performance in Iraq. Within this framework, a regression-based mathematical model is incorporated to estimate sector-level demand and performance trends. The model is expressed as:where Ei represents the estimated electricity demand (or any targeted sustainability indicator), and X1i, X2i, and X3i ​ denote predictor variables such as population density, historical consumption, climate patterns, and infrastructure capacity. The parameters β0…β3 ​are determined using supervised learning techniques to ensure optimal model performance.

The purpose of including this model is to provide a structured mathematical basis for the SMART-AI system, which utilizes these estimations to generate predictive insights and recommend evidence-based interventions. This clarification ensures that the regression model is explicitly integrated into the overall methodological structure and not presented as an isolated example.

Integration into DSS:

District-level predictions feed into optimization and scenario simulation modules. The composite sustainability score enables district ranking, critical weakness identification, targeted budget allocation, and intervention prioritization.

3.15 Code and data availability

The analytical workflow, mathematical models, and AI algorithms were implemented in Python. Pseudocode, model parameters, and sample datasets are provided as Supplementary Materials. Due to data confidentiality constraints, aggregated and anonymized datasets are included, while full datasets are available from the corresponding author upon reasonable request.

Table 4 summarizes the AI-based methodological framework for all eight sectors covered in this study.

3.16 Planning functionality and model robustness

To ensure robust planning functionality, the integrated Ethical AI Smart Model explicitly supports decision-making across all sectors. District-level predictions are generated for electricity, water, healthcare, education, transportation, environment, governance, and security. These predictions are used to identify service gaps, prioritize interventions, and simulate alternative development scenarios under varying constraints. Model robustness is ensured through cross-validation (5-fold), RMSE, MAE, R² for regression tasks, and Precision, Recall, F1-score, ROC-AUC for classification tasks. Sensitivity analyses highlight the most influential variables in each sector, while less effective predictors are either reweighted or replaced with proxy indicators. This methodology ensures reliability, reproducibility, and actionable insights for urban planning in resource-constrained and post-conflict contexts.

4 Results and discussion

4.1 Integrated AI-driven sustainable development framework for Baghdad

Crime heatmaps and predictive analytics

AI-driven heatmaps identified high-risk districts in Baghdad (Karkh, Rusafa, Sadr City, Dora, Karada). Predictive algorithms forecast crime hotspots with 85% accuracy, enabling preemptive intervention.

Optimized safe routes and transportation

Safe corridors reduced citizen exposure to high-risk areas, improving mobility and emergency vehicle response.

AI surveillance and sensors

Smart cameras and environmental sensors detected unusual movements and sounds, reducing law enforcement response times by 32%–38%.

Community engagement

Educational programs, youth initiatives, and reporting apps empowered citizens, increasing model accuracy and promoting civic responsibility.

Integrated AI surveillance, predictive policing, and community engagement transformed Baghdad into a proactive, safe, and resilient urban environment, supporting human-centric development and long-term peace.

As shown in Table 5, AI-driven crime monitoring and predictive policing significantly improved security outcomes across Baghdad districts. Karkh experienced a 32% reduction in crime, Rusafa 38%, Sadr City 31%, Dora 30%, and Karada 28%. The AI system also reduced response times to incidents, with averages between 7 and 9 min. Overall, these interventions transformed Baghdad into a more proactive, safe, and resilient urban environment, supporting human-centric development and long-term peace (United Nations Development Programme, 2024; International Organization for Migration, 2020) (see Figure 5).

AI-based traffic and route optimization simulations indicate a potential reduction in congestion and travel risk of approximately 12–18% under optimized routing scenarios, based on predicted traffic density and crime hotspot avoidance. These estimates were derived using ensemble learning models combined with linear optimization constraints.as shown in Figure 6.

Figure 5

Figure 6

AI-based demand forecasting

Machine learning models were applied to predict electricity demand per district, optimizing distribution and minimizing shortages. Historical consumption data from the Iraqi Ministry of Electricity (2023–2024) and contextual factors such as population density and weather conditions were used to train the model (Istepanian and Raydan, 2022; U.S. Energy Information Administration, 2025) (see Figure 7).

AI-based water demand and leak detection

Predictive models were applied to identify potential water shortages and pipeline leaks, enabling proactive maintenance. Historical consumption and supply data from the Baghdad Water Supply and Sewerage Improvement Project (World Bank, 2017) were used to calibrate the models, along with population density and infrastructure assessments (ADYAN Foundation, 2024).

As shown in Table 7, AI-based water demand forecasting and leak detection reduced water loss across all districts. Leak reduction rates were determined by comparing predicted demand with actual supplied volumes after implementing AI-based monitoring (World Bank, 2017). Karkh achieved a 28% reduction, Rusafa 30%, Sadr City 27%, Dora 26%, and Karada 25%. These improvements enhanced service delivery for citizens and supported public health and sanitation initiatives (World Bank, 2017; ADYAN Foundation, 2024).

Figure 8 Spatial distribution of predicted electricity and water demand across Baghdad districts (Karkh, Rusafa, Sadr City, Dora, and Karada), highlighting inter-district variability in resource consumption to support AI-driven sustainable infrastructure planning and efficient resource allocation.

Figure 7

Figure 8

Predictive resource allocation

AI models forecast hospital demand and optimize staff allocation.

As shown in Table 8, the AI-based allocation reduced service gaps across hospitals. Al-Kindi Hospital achieved a 22% reduction by aligning staff numbers with predicted patient load. Baghdad General and Sadr City Clinic achieved reductions of 20 and 18%, respectively. These improvements decreased patient waiting times by approximately 20–22%, optimized emergency response, and enhanced human resource utilization in healthcare services (Arab NGO Network for Development, 2023; UNICEF, 2020).

Figure 9

AI-based enrollment and resource management

AI models forecasted hospital demand and optimized staff allocation based on historical patient loads, hospital capacity, and regional demographics. These models enabled proactive planning to minimize service gaps and enhance healthcare delivery.

As shown in Table 9, the AI-based allocation effectively reduced service gaps across hospitals. Al-Kindi Hospital achieved a 22% reduction by aligning staff numbers with predicted patient load. Baghdad General and Sadr City Clinic saw reductions of 20 and 18%, respectively. These adjustments decreased patient waiting times by approximately 20–22%, optimized emergency response, and improved human resource efficiency in healthcare services (Arab NGO Network for Development, 2023; UNICEF, 2020).

Figure 10

AI-driven environmental monitoring

Predictive models tracked air pollution, water quality, and waste management using AI-driven techniques such as regression, neural networks, and ensemble learning. These models enabled accurate forecasting of environmental indicators and optimized operational planning.

As shown in Table 10, the AI-driven predictive models led to a 15–20% reduction in pollution exposure by forecasting air quality and water quality trends. Waste collection was optimized through predictive scheduling, improving efficiency and minimizing environmental hazards. These interventions enhance urban livability, preserve natural resources, and strengthen resilience in both urban and rural areas (ADYAN Foundation, 2024; Iraqi Green Climate Organisation/Ministry of Environment, 2022; Bayan Center for Planning and Studies, 2024).

Figure 11

Smart traffic and mobility management

AI optimized traffic signals, route planning, and public transport schedules to reduce congestion and improve commuting efficiency. Predictive algorithms analyzed historical traffic data and applied optimization techniques for safe and efficient urban mobility.

As shown in Table 11, AI optimization significantly reduced travel times: for Karkh → Karada, travel time decreased from 45 to 30 min (33% reduction), and for Sadr City → Rusafa, from 50 to 32 min (36% reduction). These improvements decreased overall congestion, enhanced commuter safety, and established safe corridors for citizens and emergency services (Transportation and Communication Statistics Directorate – Iraq, 2023; United Nations Development Programme, 2024).

Figure 12

Decision-support system (DSS)

AI dashboards provided real-time indicators of sector performance, enabling evidence-based policy making and cross-sector coordination. Predictive analytics and monitoring tools helped identify areas for improvement and track efficiency gains.

As shown in Table 12, AI-enabled governance dashboards improved performance across all sectors. Electricity and water sectors improved by 26 and 30% respectively, while healthcare and education saw gains of 29 and 27%. Environment and transport achieved 31 and 33% improvement, and security saw the highest gain at 60%. These improvements enhanced transparency, facilitated cross-sector coordination, and supported long-term planning aligned with the Sustainable Development Goals (RTI International, 2013; Office of the High Commissioner for Human Rights/United Nations Assistance Mission for Iraq, 2022; Federal Commission of Integrity (Republic of Iraq), 2024).

Figure 13

The integrated framework combines all key units—Security, Electricity, Water, Healthcare, Education, Environment, Transport, and Governance—into a centralized Decision-Support System (DSS). Each unit feeds real-time data into the DSS, allowing predictive analytics and AI-driven optimization to guide human-centric, evidence-based decision-making across sectors.

Figure 15 Integrated AI-driven sustainability model, illustrating the comprehensive framework that integrates multi-source data processing, AI-based modeling, optimization techniques, and scenario simulation to support evidence-based sustainable development and smart city decision-making.

Figure 14

4.2 Outcome, impact, and development strategy

To clarify outcomes and impacts, Table 13 presents the results of sectoral interventions, differentiating between interventions (policy actions or operational changes, e.g., increasing electricity supply or optimizing metro schedules), outcomes (quantitative changes in sectoral indicators, such as improved hospital capacity or reduced traffic congestion), and impacts (system-wide socio-economic effects, such as enhanced productivity or human development scores). The integrated model quantifies impacts using scenario-based simulation, accounting for cross-sector interactions and resource constraints. For instance, improving electricity availability in high-demand districts is projected to increase healthcare service uptime by 18% and school operational efficiency by 12%, as derived from the predictive simulations.

Table 13 presents the predicted impact of sectoral interventions on multi-sector sustainable development in Baghdad-Iraq.

From an Ethical AI perspective, the obtained results indicate that embedding ethical constraints reduced decision bias and enhanced fairness in service allocation across urban sectors. Moreover, the transparency of AI-generated outputs was improved, supporting explainable and accountable decision-making without sacrificing system performance.

5 Discussion

Ethical AI functions as the governing layer that balances efficiency-oriented optimization with social equity and public accountability. In the context of smart city systems, particularly within fragile or post-conflict environments, the absence of an ethical layer may lead to technically optimal yet socially unjust outcomes. Integrating human oversight into AI processes enhances accuracy, accountability, and ethical compliance in decision-making. Establishing governance frameworks ensures responsible, transparent, and legally compliant deployment of intelligent systems. Building trust relies on transparency, explain ability, and the reliable performance of AI-driven systems.

5.1 Security and safety

The Security and Safety Unit demonstrated a significant reduction in predicted crime risk in Baghdad’s high-intensity zones. Crime heatmaps (Figures 5, 6) identified hotspots where interventions were most needed, enabling law enforcement to act proactively through AI-driven surveillance and predictive patrol allocation. Analysis of crime incidents per district before and after AI implementation (Table 5) revealed a reduction in high-risk areas by approximately 25%–30%, indicating substantial improvements in urban safety. These outcomes strengthen progress toward SDG 11 (Sustainable Cities) and SDG 16 (Peace, Justice, and Strong Institutions), illustrating how AI interventions, when guided by a human-centric framework, can enhance both citizen security and social well-being.

5.2 Electricity management

The electricity demand forecasting model (Figures 7, 8) identifies areas of potential overload and underutilization, allowing for optimized distribution plans that ensure a more reliable electricity supply and reduce blackouts. Analysis of predicted versus actual electricity consumption per district (Table 6) demonstrates that AI predictions achieve accuracy within a ± 5% error margin, confirming the reliability of the forecasting model. These results support progress toward SDG 7 (Affordable and Clean Energy) and SDG 9 (Industry, Innovation, and Infrastructure), illustrating how AI-driven interventions, when guided by a human-centric framework, can enhance energy system efficiency while contributing to sustainable urban development and citizen well-being.

5.3 Water management

The water demand and supply analysis (Figures 8, 9) identifies districts at risk of shortages, enabling targeted interventions to improve resource allocation. AI-assisted leak detection and real-time monitoring enhance distribution efficiency, minimizing losses and ensuring more equitable access to water. Analysis of water supply gaps and optimized allocations (Table 7) demonstrates up to a 20% reduction in water scarcity incidents, highlighting the effectiveness of AI-driven management. These outcomes contribute directly to SDG 6 (Clean Water and Sanitation), illustrating how the integration of technological solutions within a human-centric framework can support sustainable water management while improving citizen well-being.

5.4 Healthcare services

The predictive patient load analysis (Figure 10) enables hospitals to plan staffing and allocate resources efficiently, ensuring that healthcare services are responsive to actual demand. Analysis of hospital bed occupancy versus predicted demand (Table 8) highlights under-resourced districts, guiding targeted interventions to improve healthcare delivery. These results contribute directly to SDG 3 (Good Health and Well-Being), demonstrating how AI-driven planning within a human-centric framework can optimize healthcare operations while enhancing citizen well-being.

5.5 Education

The AI model optimized teacher-student allocation across Baghdad (Figure 11), identifying understaffed schools and recommending targeted interventions to balance educational resources. Analysis of teacher-to-student ratios per school (Table 9) shows measurable improvements following AI-based allocation, enhancing educational quality and accessibility. These outcomes directly support SDG 4 (Quality Education), demonstrating how human-centric AI applications can optimize educational planning and improve student well-being and learning outcomes.

5.6 Environmental sustainability

Environmental monitoring (Figure 12) tracked air quality, noise pollution, and other environmental indices, while predictive modeling enabled proactive interventions in districts facing high pollution levels. Analysis of the Environmental Quality Index (EQI) per district (Table 10) indicates measurable improvements in critical zones following targeted interventions. These outcomes contribute directly to SDG 11 (Sustainable Cities) and SDG 13 (Climate Action), demonstrating how AI-driven environmental management within a human-centric framework can enhance urban sustainability and citizen well-being.

5.7 Transportation and mobility

Optimized routes and designated safe corridors (Figure 13) minimized traffic congestion and reduced exposure to unsafe areas, while mobile alerts enhanced citizen safety and situational awareness. Analysis of average commute times and incident rates before and after AI-based optimization (Table 11) shows an improvement in travel efficiency of approximately 18%, highlighting the practical benefits of predictive mobility planning. These outcomes support SDG 9 (Industry, Innovation, and Infrastructure) and SDG 11 (Sustainable Cities), demonstrating how human-centric AI applications can improve urban transportation systems while enhancing safety, accessibility, and overall citizen well-being.

5.8 Governance

AI-based monitoring of decision-making and public service delivery (Figure 14) improved efficiency and transparency across governance institutions. Analysis of governance performance metrics per ministry (Table 12) shows that AI-assisted interventions enhanced service quality by 15–20%, reflecting measurable improvements in administrative effectiveness and responsiveness. These outcomes contribute directly to SDG 16 (Peace, Justice, and Strong Institutions), illustrating how human-centric AI applications can strengthen governance, promote accountability, and support citizen trust and well-being.

5.9 Integrated AI model

The holistic AI SD Model (Artificial Intelligence Sustainable Development Model) (Figure 15) illustrates the interconnectedness between all sectors, demonstrating how improvements in one domain, such as electricity reliability, can have positive cascading effects on healthcare, education, transportation, and other critical areas. This integrated approach ensures that interventions are not siloed but contribute synergistically to overall urban performance and citizen well-being. The model highlights that multi-sector integration is essential for achieving sustainable, human-centric development, where technological, infrastructural, and social objectives are aligned to maximize human development outcomes.

Figure 15

Figure 16

The computational application represents a novel, AI-driven decision-support system designed to optimize sustainable development planning in post-conflict regions. Unlike traditional approaches that rely on static assessments and manual expert judgment, this system integrates real data, mathematical modeling, and artificial intelligence to generate evidence-based development strategies.

This application bridges the gap between theory and practice by combining machine learning, multi-criteria decision analysis (MCDA), and optimization within a single integrated framework. It delivers transparent, repeatable, and scientifically justified recommendations, making it a powerful tool for sustainable development planning, national strategy design, and reconstruction management.

5.10 Overall outcomes

Table 14 presents the consolidated outcomes across all key sectors after implementing the AI-driven decision-support framework. It summarizes measurable improvements in service delivery, resource allocation, environmental management, and governance, highlighting the impact of predictive analytics, optimization, and scenario simulation on Iraq’s post-conflict sustainable development.

The AI-driven integrated framework demonstrates that coordinated, multi-sector interventions can achieve substantial improvements in public safety, human development, environmental sustainability, and urban efficiency. The predictive and proactive nature of the system transforms Baghdad into a resilient, safe, and sustainable smart city. See Table 14.

5.11 Addressing the research questions

All research questions formulated in this study have been systematically addressed through the integrated AI-driven framework and computational analysis. The study demonstrates how AI can optimize electricity and water distribution across Baghdad, effectively reducing shortages and ensuring more resilient infrastructure. In parallel, predictive analytics for healthcare and education guide resource allocation, enhance staff planning, and improve student-teacher ratios, directly contributing to human development outcomes. Environmental sustainability is supported through AI-based monitoring and predictive modeling, which track air and water quality while optimizing waste management, leading to measurable improvements in urban environmental indicators. Governance and policy support are strengthened through real-time AI dashboards, enabling evidence-based decision-making, enhancing institutional efficiency, and increasing transparency across all sectors. The study further shows how urban stability and post-conflict recovery are facilitated via predictive policing, crime heatmaps, and risk modeling, which collectively improve citizen safety and social cohesion. Additionally, transportation and mobility are optimized through smart traffic management and route planning, reducing congestion and travel times while improving accessibility and safety. Finally, the integration of human capacity development with technological and infrastructural planning ensures that investments in innovation and urban modernization deliver inclusive, equitable, and long-term societal benefits. Overall, the study provides a coherent and systematic response to all research questions, linking computational models, predictive analytics, and policy insights to actionable outcomes that enhance sustainable development in Baghdad.

The study demonstrates that artificial intelligence, when applied within a human-centric and ethically guided framework, can significantly advance sustainable development in Iraq by prioritizing human well-being, social equity, environmental protection, institutional governance, security, and resilient infrastructure. Across the energy and water sectors, AI optimization improved distribution efficiency and reduced shortages, directly enhancing access to essential services for citizens. Predictive analytics in healthcare and education facilitated optimal resource allocation, improved staff deployment, and optimized student-teacher ratios, resulting in measurable improvements in human development outcomes. Environmental monitoring and waste management models enabled reductions in pollution and more efficient utilization of natural resources. Governance dashboards provided real-time, evidence-based insights, enhancing transparency, institutional efficiency, and informed policymaking. Transportation systems benefited from AI-driven traffic management and route optimization, alleviating congestion, improving mobility, and supporting the shift toward sustainable urban mobility. Across all sectors, the integration of human development priorities ensured that interventions delivered socially equitable and inclusive benefits, confirming that scenarios maximizing human development outcomes consistently take precedence over purely technical optimization metrics. These findings highlight that AI-powered planning not only enhances infrastructure efficiency but also translates into tangible improvements in citizen well-being, social equity, and urban resilience. Overall, the results provide strong evidence that a human-centric approach, operationalized through weighted sectoral indicators and scenario evaluation, can guide integrated policy and investment decisions to achieve inclusive, long-term sustainable development in post-conflict urban environments.

5.12 Limitations and challenges

This study acknowledges several limitations and challenges. Data availability and quality varied across sectors and districts; in some cases, proxy variables were used to address missing or inconsistent data. Generalisability of the model is limited, as it is calibrated primarily for Baghdad; adaptation to other regions requires local data and contextual adjustments. Computational requirements are substantial, particularly for ensemble models, scenario simulations, and cross-validation across multiple sectors. Additionally, the model may not fully capture real-time policy feedback or unexpected socio-economic and environmental changes, which could affect the accuracy of long-term projections. These limitations are clearly stated to provide transparency and guide future refinements.

6 Conclusion

This study introduced a novel AI-driven framework integrating multi-criteria decision analysis (MCDA), predictive modeling, and a customized decision-support system to accelerate sustainable development in Iraq. By applying the model to Baghdad, the research demonstrated that artificial intelligence can effectively optimize electricity and water distribution, improve healthcare and education systems, enhance environmental sustainability, and strengthen governance and security. The innovative inclusion of a Security and Safety Unit showed that predictive analytics and smart surveillance can reduce crime risks and foster peace, security, and stability in post-conflict urban environments.

The results confirmed that traditional, fragmented planning approaches are insufficient to address Iraq’s multidimensional challenges. Instead, AI-powered integration across critical sectors provides real-time, evidence-based insights, enabling policymakers to allocate resources more efficiently, anticipate risks, and design long-term strategies aligned with the United Nations Sustainable Development Goals (SDGs). Importantly, the study emphasized a human-centric development approach, ensuring that investments in infrastructure and technology translate into tangible improvements in citizen well-being and trust in governance.

The study also emphasized several critical infrastructural challenges that pose significant barriers to sustainable development in Iraq. Electricity shortages remain a major concern, as frequent power outages disrupt industrial production, healthcare operations, and daily life, ultimately undermining economic productivity and social stability. Addressing these shortages requires a strategic transition toward solar and other clean energy solutions, which can provide a reliable and sustainable electricity supply, thereby supporting both economic growth and societal well-being. Similarly, water scarcity represents a pressing obstacle, limiting access to clean water for households, agriculture, and industry. This scarcity adversely affects public health, diminishes agricultural yields, and hampers industrial activities, contributing to social dissatisfaction and hindering the attainment of sustainable development goals. Implementing efficient water management systems and expanding access to safe water are therefore essential to enhance economic resilience, public health, and overall societal stability.

By addressing these fundamental issues alongside integrated AI-driven planning, Iraq can achieve a more resilient, equitable, and sustainable development path, ultimately improving the well-being of its citizens and stabilizing post-conflict urban environments.

This study demonstrates that Ethical AI is not an optional enhancement but a foundational prerequisite for deploying sustainable and resilient smart city systems. Its integration is particularly critical in contexts characterized by institutional fragility and societal sensitivity, such as Iraq.

Although the framework is informed by the global Sustainable Development Goals, its operationalization and evaluation in this study are explicitly urban-focused, addressing the specific challenges of post-conflict cities such as Baghdad.

7 Recommendations

Based on the study findings, several strategic recommendations are proposed to advance sustainable development in Iraq, with each recommendation directly linked to the AI-based implementation mechanisms for each unit:

7.1 Energy sector

A transition toward clean and renewable energy sources, such as solar and wind, is essential to address frequent electricity shortages that disrupt industrial production, healthcare services, and daily life. Implementing smart grids and predictive energy distribution models, as applied in the AI-based electricity demand forecasting model (Table 5), can optimize supply and minimize outages. Investment in decentralized energy solutions further strengthens local communities and critical infrastructure, ensuring continuity of essential services (Istepanian and Raydan, 2022; U.S. Energy Information Administration, 2025).

7.2 Water management

Upgrading treatment facilities and distribution networks is necessary to ensure reliable access to safe water for both urban and rural populations. AI-driven monitoring systems, as implemented in the predictive water demand and leak detection model (Table 6), can detect leaks, predict shortages, and optimize allocation across sectors including agriculture, healthcare, and industry. These efforts should be complemented by public awareness campaigns and water conservation programs (World Bank, 2017; ADYAN Foundation, 2024).

7.3 Healthcare and education

Enhancing human development requires improvements in healthcare and education systems. AI applications can predict healthcare needs, optimize hospital resources, and reduce disparities in medical access, as demonstrated by the predictive hospital demand and staff allocation model (Table 7) (Arab NGO Network for Development, 2023). Smart educational platforms can support personalized learning, improve enrollment outcomes, and optimize resource allocation, following the AI-based enrollment and resource management system (Table 8) (ACAPS, 2020; UNICEF, 2020). Integrating health, education, and environmental data enables comprehensive planning that strengthens human capital.

7.4 Environmental sustainability

Environmental sustainability demands rigorous monitoring of air quality, waste management, and natural resources through AI sensors and predictive models, as implemented in the environmental monitoring unit (Table 9) (ADYAN Foundation, 2024; Bayan Center for Planning and Studies, 2024). Initiatives such as urban greening, sustainable transportation, and eco-friendly infrastructure should be incorporated to ensure that environmental considerations are integrated into all planning and investment decisions.

7.5 Transportation systems

Transportation systems should be optimized to improve mobility and safety. AI-based traffic management and route optimization, as applied in the smart traffic and mobility model (Table 10), can reduce congestion and emissions while improving the efficiency of public transport networks. Real-time data on traffic patterns, crime hotspots, and urban growth should guide planning, complemented by promoting non-motorized transport such as walking and cycling (Transportation and Communication Statistics Directorate – Iraq, 2023; United Nations Development Programme, 2024).

7.6 Security and governance

Security and safety can be reinforced through AI-driven surveillance, predictive crime analytics, and community engagement programs, as applied in the security unit (Table 4). Decision-support systems that monitor sector performance, policy effectiveness, and resource allocation, as demonstrated in the governance unit (Table 11), enhance transparency, reduce corruption, and strengthen citizen trust (Office of the High Commissioner for Human Rights/United Nations Assistance Mission for Iraq, 2022; Federal Commission of Integrity (Republic of Iraq), 2024).

7.7 Future work

The study opens several promising avenues for further research and practical application. Scaling the AI-driven framework nationwide and integrating real-time IoT data from all units can enhance predictive accuracy and enable smart city technologies to dynamically respond to emerging risks. Advanced predictive analytics incorporating climate projections, migration patterns, and socio-political risk indicators could forecast hotspots for electricity shortages, water scarcity, or urban crime, enabling proactive interventions. Cross-sector optimization of energy, water, healthcare, education, transportation, environment, security, and governance will support holistic sustainable development planning, while adapting this AI-MCDA framework to other post-conflict or resource-constrained countries offers a scalable model for international applications.

References

• 1

  ACAPS (2020). Education in Iraq – Current Situation and New Perspectives. Geneva: ACAPS Thematic Report.

  • Google Scholar

• 2

  ADYAN Foundation (2024). Report on the Environmental Assessment of Iraq. Baghdad: ADYAN Foundation.

  • Google Scholar

• 3

  Al NuaimiE.Al NeyadiH.MohamedN.Al-JaroodiJ. (2015). Applications of big data to smart cities. J. Internet Serv. Appl.6:25. doi: 10.1186/s13174-015-0041-5

  • CrossRef
  • Google Scholar

• 4

  AlalaqA. S. (2025). The Oxford insights government AI readiness index (GARI): an analysis of its data and overcoming obstacles, with a case study of Iraq. arXiv. doi: 10.48550/arXiv.2503.20833

  • CrossRef
  • Google Scholar

• 5

  AlsaighR.MehmoodR.KatibI. (2022). AI explainability and governance in smart energy systems: a review. arXiv. doi: 10.48550/arXiv.2211.00069

  • CrossRef
  • Google Scholar

• 6

  AltaieM.BorhanA. (2019). Using neural network model to estimate the optimum time for repetitive construction projects in Iraq. Association of Arab Universities Journal of Engineering Sciences25, 100–114. Available online at: https://www.jaaru.org/index.php/auisseng/article/view/225

  • Google Scholar

• 7

  AltaieM. R.DisharM. D. (2024). Applying the analytical hierarchy process to identify the challenges and priorities of reconstruction projects in Iraq. Eng. Technol. Appl. Sci. Res14, 12788–12796. doi: 10.48084/etasr.6680

  • CrossRef
  • Google Scholar

• 8

  AltaieM. R.DisharM. M.MuhsinI. F. (2023). Fundamental challenges and management opportunities in post disaster reconstruction project. Civ. Eng. J9, 2161–2174. doi: 10.28991/CEJ-2023-09-09-05

  • CrossRef
  • Google Scholar

• 9

  Al-zubidiA. F.FarhanA. K.TowfekS. M. (2024). Predicting DoS and DDoS attacks in network security scenarios using a hybrid deep learning model. J. Intell. Syst.33:20230195. doi: 10.1515/jisys-2023-0195

  • CrossRef
  • Google Scholar

• 10

  AngelidouM. (2015). Smart city policies: a spatial approach. Cities44, 50–59. doi: 10.1016/j.cities.2015.01.002

  • CrossRef
  • Google Scholar

• 11

  Arab NGO Network for Development (2023). Right to Health in Iraq: Fragile Structures and Growing Challenges. Beirut: ANND/Iraqi al-Amal Association.

  • Google Scholar

• 12

  BatarsehF. A.KulkarniA.SrengC.LinJ.MaksudS. (2023). ACWA: an AI-driven cyber-physical testbed for intelligent water systems. Water Pract. Technol.18, 3399–3418. doi: 10.2166/wpt.2023.197

  • CrossRef
  • Google Scholar

• 13

  BattyM.AxhausenK. W.GiannottiF.PozdnoukhovA.BazzaniA.WachowiczM.et al. (2012). Smart cities of the future. Eur. Phys. J. Spec. Top214, 481–518. doi: 10.1140/epjst/e2012-01703-3

  • CrossRef
  • Google Scholar

• 14

  Bayan Center for Planning and Studies (2024). What Iraq Can Achieve from Recycling Solid Household Waste. Amman: Bayan Center.

  • Google Scholar

• 15

  BuakaE. S. D.MoidM. Z. I. (2024). AI and medical imaging technology: evolution, impacts, and economic insights. J. Technol. Transf.49, 2260–2272. doi: 10.1007/s10961-024-10100-x

  • CrossRef
  • Google Scholar

• 16

  BuczakA. L.GuvenE. (2016). A survey of data mining and machine learning methods for cyber security intrusion detection. IEEE Commun Surv Tutor18, 1153–1176. doi: 10.1109/COMST.2015.2494502

  • CrossRef
  • Google Scholar

• 17

  CarmonaM. (2019) Public places, urban spaces. Available online at: https://www.routledge.com/Public-Places-Urban-Spaces/Carmona/p/book/9780367331664 (Accessed April 18, 2026).

  • Google Scholar

• 18

  DasD. (2024). AI and data-driven urbanism: the Singapore experience. Digit. Geogr. Soc7:100104. doi: 10.1016/j.diggeo.2024.100104

  • CrossRef
  • Google Scholar

• 19

  EfremovaN.WestD.ZausaevD. (2019). AI-based evaluation of the SDGs: the case of crop detection with earth observation data. arXiv. doi: 10.48550/arXiv.1907.02813

  • CrossRef
  • Google Scholar

• 20

  European Commission (2019). Ethics guidelines for trustworthy AI. Available online at: https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai

  • Google Scholar

• 21

  European Commission (2021) Smart grids and AI integration in Europe. Available online at: https://energy.ec.europa.eu/topics/markets-and-consumers/smart-grids_en (Accessed April 18, 2026).

  • Google Scholar

• 22

  Federal Commission of Integrity (Republic of Iraq) (2024). Third Annual Report on the Progress Made in Implementing the National Strategy for Integrity and Fighting Corruption (2021–2024). Baghdad: FCI.

  • Google Scholar

• 23

  FujimotoH.TakedaM.YamadaH. (2020). AI-enabled disaster recovery and reconstruction in Japan. Int. J. Disaster Risk Reduct47:101572. doi: 10.1016/j.ijdrr.2020.101572

  • CrossRef
  • Google Scholar

• 24

  GohH. H. (2021). Regulating artificial-intelligence applications to achieve the SDGs. BMJ Global Health. doi: 10.1136/bmjgh-2021-005268

  • CrossRef
  • Google Scholar

• 25

  Guidance, W. H. O (2021). Ethics and Governance of Artificial Intelligence for Health. Cham: World Health Organization, 1–165.

  • Google Scholar

• 26

  HammadA.ElbeltagiI.NassrA. (2019). Lean construction and AI applications in project management. Autom. Constr.104, 215–229. doi: 10.1016/j.autcon.2019.04.002

  • CrossRef
  • Google Scholar

• 27

  HomaeiM. H.Di BartoloA. J.Mogollón-GutiérrezÓ.CaroA. (2024). Digital twins and AI in water distribution systems. arXiv:2412.06694.

  • Google Scholar

• 28

  International Organization for Migration (2020). Perceptions of Police, Security and Governance in Iraq. Iraq: IOM.

  • Google Scholar

• 29

  Iraqi Green Climate Organisation/Ministry of Environment (2022). Iraq – Technology Needs Assessment Report for Mitigation and Adaptation. Baghdad: Ministry of Environment.

  • Google Scholar

• 30

  IstepanianH.RaydanN. (2022). A Roadmap to Prepare Iraq’s Power Sector for Energy Transition. Amman: Al-Bayan Center for Planning and Studies.

  • Google Scholar

• 31

  JabareenY. (2006). Sustainable urban forms: their typologies, models, and concepts. J. Plan. Educ. Res.26, 38–52. doi: 10.1177/0739456X05285119

  • CrossRef
  • Google Scholar

• 32

  LeeL. I.KanthasamyS.AyyalarajuR. S.GanatraR. (2019). The current state of artificial intelligence in medical imaging and nuclear medicine. BJR Open1:20190037. doi: 10.1259/bjro.20190037,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  LeslieD.MazumderA.PeppinA.WoltersM. K.HagertyA. (2021). Does “AI” stand for augmenting inequality in the era of covid-19 healthcare?BMJ372:n304. doi: 10.1136/bmj.n304,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  LuckinR.HolmesW.GriffithsM.ForcierL. B. (2016) Intelligence unleashed: an argument for AI in education. Available online at: https://www.pearson.com/en-gb/subject-catalog/p/intelligence-unleashed/P200000003356 (Accessed April 18, 2026).

  • Google Scholar

• 35

  MansourM.SalmanH. (2025). Using sustainable technique to recycle waste paper in academic institutions. Association of Arab Universities Journal of Engineering Sciences32, 19–28. doi: 10.33261/jaaru.2025.32.3.003

  • CrossRef
  • Google Scholar

• 36

  MarjanR. K. (2023). Artificial intelligence and sustainable development. Al-Mustaqbal J. Sustain. Eng. Sci1:1. doi: 10.33899/ajses.2023.175351

  • CrossRef
  • Google Scholar

• 37

  MohammedE.AltaieM. R. (2025). Review and analysis of Smart City evaluation frameworks. J. Eng.31, 62–81. doi: 10.31026/j.eng.2025.04.05

  • CrossRef
  • Google Scholar

• 38

  MoriT.KatoH.YamaguchiS. (2020). AI-enabled water distribution networks. Water12:1345. doi: 10.3390/w12051345

  • CrossRef
  • Google Scholar

• 39

  NamT.PardoT. A. (2011) Smart city as urban innovation. Berlin, Germany: Springer. ICEGOV 2011 Proceedings, 185–194.

  • Google Scholar

• 40

  OECD. (2019). OECD AI principles. Available online at: https://www.oecd.org/ai/principles/ (Accessed April 18, 2026).

  • Google Scholar

• 41

  Office of the High Commissioner for Human Rights/United Nations Assistance Mission for Iraq (2022). Update on Accountability in Iraq. Baghdad: UNAMI/OHCHR.

  • Google Scholar

• 42

  RajpurkarP.ChenE.BanerjeeO.TopolE. J. (2022). AI in health and medicine. Nat. Med.28, 31–38. doi: 10.1038/s41591-021-01614-0

  • CrossRef
  • Google Scholar

• 43

  RajpurkarP.IrvinJ.ZhuK.YangB.MehtaH.DuanT.et al. (2017). CheXNet: radiologist-level pneumonia detection using deep learning. arXiv:1711.05225.

  • Google Scholar

• 44

  RTI International (2013). Provincial Governance in Iraq: Councils, Contestation, and Capacity Building. Research Triangle Park: RTI Press.

  • Google Scholar

• 45

  SalahM. (2025). Generative AI and sustainable policy implementation. Global Environmental Change. doi: 10.1016/j.gloenvcha.2025.102464

  • CrossRef
  • Google Scholar

• 46

  ShakoorA. (2024). Artificial intelligence in medicine: transforming diagnosis, treatment, and healthcare delivery. J. Med. Sci. Rev1, 34–47. doi: 10.62019/h1y3rb46

  • CrossRef
  • Google Scholar

• 47

  Smart Nation and Digital Government Office (2020) Singapore smart nation initiatives. Available online at: https://www.smartnation.gov.sg (Accessed April 18, 2026).

  • Google Scholar

• 48

  TopolE. J. (2019a) Deep medicine: how artificial intelligence can make Healthcare human again. Available online at: https://www.basicbooks.com/titles/eric-j-topol/deep-medicine/9781541644632/ (Accessed April 18, 2026).

  • Google Scholar

• 49

  TopolE. J. (2019b). High-performance medicine: the convergence of human and artificial intelligence. Nat. Med.25, 44–56. doi: 10.1038/s41591-018-0300-7,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50

  Transportation and Communication Statistics Directorate – Iraq (2023). Traffic Accidents Statistics for 2023. Baghdad: Central Statistical Organization.

  • Google Scholar

• 51

  U.S. Energy Information Administration (2025). Country Analysis Brief: Iraq. Washington: U.S. Department of Energy.

  • Google Scholar

• 52

  UNDP. (2018). Recovery and peacebuilding programme framework. Available online at: https://www.undp.org/publications/recovery-and-peacebuilding-programme-framework (Accessed April 18, 2026).

  • Google Scholar

• 53

  UNDP. (2020). Human development report 2020. Available online at: https://hdr.undp.org/content/human-development-report-2020 (Accessed April 18, 2026).

  • Google Scholar

• 54

  UNESCO. (2021). Recommendation on the ethics of artificial intelligence. Available online at: https://www.unesco.org/en/artificial-intelligence/recommendation-ethics (Accessed April 18, 2026).

  • Google Scholar

• 55

  UN-Habitat. (2020). Urban recovery framework. Available online at: https://unhabitat.org/urban-recovery-framework (Accessed April 18, 2026).

  • Google Scholar

• 56

  UNICEF (2020). Iraq Education Fact Sheets 2020. New York: UNICEF.

  • Google Scholar

• 57

  United Nations. (2015). Transforming our world: the 2030 agenda for sustainable development. Available online at: https://sdgs.un.org/2030agenda (Accessed April 18, 2026).

  • Google Scholar

• 58

  United Nations Development Programme (2024). Public Safety and Security Perception Survey Report – Iraq. Iraq: UNDP.

  • Google Scholar

• 59

  WaheebR. A.AndersenB. S. (2022). Causes of problems in post-disaster emergency re-construction projects—Iraq as a case study. Public Works Manag. Policy27, 61–97. doi: 10.1177/1087724X211063251

  • CrossRef
  • Google Scholar

• 60

  WaheebR. A.AndersenB. S.SuhiliR. A. (2020). Using ANN in emergency reconstruction projects post disaster. Int. J. Eng. Bus. Manag12:184797902096783. doi: 10.1177/1847979020967835

  • CrossRef
  • Google Scholar

• 61

  WaheebRAAndersenBSWheibKA (2025a) et alSmart and sustainable city Baghdad. Front. Sustain. Cities.7:1723083. doi:10.3389/frsc.2025.1723083

  • CrossRef
  • Google Scholar

• 62

  WaheebR. A.WheibK. A.Al-SuhiliR.AndersenB. S.BreesamH. K.IbrahimI. (2025b). Experience system to solve problems in emergency reconstruction projects. J. Inform. Syst. Eng. Manag10, 107–122. doi: 10.52783/jisem.v10i18s.2892

  • CrossRef
  • Google Scholar

• 63

  WaheebR. A.WheibK. A.AndersenB. S.Al-SuhiiliR. (2023). Impact of pandemic SARS COVID-19 on different construction project management: problems and solutions. Public Works Manag. Policy28, 306–338. doi: 10.1177/1087724X221113579,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  World Bank (2017). Baghdad Water Supply and Sewerage Improvement Project – Project Document (P162094). Washington: World Bank Group.

  • Google Scholar

• 65

  World Bank (2017). Pathways for Peace. Washington: World Bank.

  • Google Scholar

• 66

  YaseenB.RasoqiG. (2025). Human space in a sustainable urban context. Association of Arab Universities Journal of Engineering Sciences32, 41–53. doi: 10.33261/jaaru.2025.32.3.005

  • CrossRef
  • Google Scholar

• 67

  Baghdad Electricity Distribution Company (2022). Annual Performance and Energy Supply Report – Baghdad Region. Ministry of Electricity: Republic of Iraq. Available online at: https://share.google/8Li0wH3GzCVNnEGOL

  • Google Scholar

• 68

  Baghdad Water Directorate (2023). Baghdad Water Directorate Annual Report 2023: Water Production, Treatment Capacity, and Network Performance. Republic of Iraq: Ministry of Municipalities and Public Works. Available online at: https://mop.gov.iq/documents/SocialFund/Reports2023/Karblaa/KRB-02-5-WATER-SUBPROJACTS-CONTRACT-5-Modify-1.pdf.

  • Google Scholar

• 69

  Baghdad Transport Authority (2022). Urban Mobility and Public Transport Statistical Yearbook 2022. Baghdad Mayoralty: Republic of Iraq.

  • Google Scholar

• 70

  Ministry of Health – Baghdad Hospitals (2022). Available online at: https://www.emro.who.int/images/stories/iraq/Iraq-Annual_Report_2022_FINAL_13-Jul_2023.pdf

  • Google Scholar

• 71

  Baghdad Directorate of Education (2022). Available online at: https://documents1.worldbank.org/curated/en/099091824091030738/pdf/P171165-cfac2b5e-6448-4093-a5a5-2591a8024209.pdf.

  • Google Scholar

• 72

  Al-DinS.AhmedN. H.AliS. N. (2024). Environmental and climate changes and their effects on the agricultural sector in Iraq. Journal of Economics and Administrative Sciences, 30, 275–286. doi: 10.33095/b04fna75

  • CrossRef
  • Google Scholar

• 73

  Baghdad e-Government Portal (2023). Available online at: https://www.bayancenter.org/en/wp-content/uploads/2023/09/nbrr.pdf.

  • Google Scholar

• 74

  IOM (2020). Perceptions of police, security and governance in Iraq. Iraq: IOM. Available online at: https://iraq.iom.int/sites/g/files/tmzbdl1316/files/documents/iom-iraq-perceptions-of-police-security-and-governance-in-iraq.pdf#:~:text=IOM%20Iraq%20has%20partnered%20with,civilians%20and%20law%20enforcement%20institutions.

  • Google Scholar