Extended Abstract

Introduction

The accelerating integration of Artificial Intelligence (AI) into urban management has redefined the dynamics of service delivery, crisis response, and citizen–government relations. As a transformative technology, AI introduces predictive analytics, automation, and real-time decision-making that can enhance the efficiency of public services, optimize resource use, and improve resilience across urban systems. At the same time, AI raises major concerns related to privacy protection, ethical accountability, algorithmic bias, workforce displacement, and overreliance on automated processes. The coexistence of both opportunities and risks makes public perception a decisive determinant for the pace and scope of AI adoption in cities.

This study positions the debate within the context of Mashhad, Iran’s second-largest metropolis, which is distinguished by its high population density, significant infrastructural challenges, and ambitious smart city agenda. Mashhad has recently invested in digital infrastructure, e-governance, and fiber optic expansion, creating a fertile ground for the deployment of AI-driven services. Nonetheless, while technological capacity is steadily advancing, the long-term sustainability and legitimacy of AI adoption in urban governance depend on citizen trust, acceptance, and recognition of tangible benefits.

The study employs theoretical frameworks such as the Technology Acceptance Model (TAM) and ethical AI principles (AI4People) to conceptualize citizen perceptions. These are captured across four analytical dimensions: (1) perceived threats, including privacy violations and potential misuse; (2) comfort and trust, reflecting institutional credibility and confidence in AI providers; (3) applications of AI in urban services, such as transportation, safety, and resource optimization; and (4) AI’s role in crisis management, particularly in disaster preparedness and emergency response. By exploring these dimensions, the study illuminates how citizens negotiate between optimism about improved efficiency and caution regarding societal risks.



Methodology

The research utilized a quantitative, survey-based approach, targeting residents of Mashhad as the study population. Applying Cochran’s formula for indeterminate populations, the estimated sample size was 385, of which 380 valid responses were collected in 2024. Participants were 18 years and older, and the survey was distributed via online questionnaires.

The instrument was developed through literature review and validation by urban studies experts. Structured items measured the four constructs: perceived threats, comfort/trust, urban service applications, and AI roles in crisis management. A 7-point Likert scale (from strong disagreement to strong agreement) was employed. The questionnaire also included demographic variables (age, gender, education, employment, income) and items on self-assessed AI knowledge and experience.

Reliability testing demonstrated strong internal consistency, with Cronbach’s alpha values exceeding 0.80. Validity was confirmed through exploratory factor analysis (EFA) in SPSS, supported by a Kaiser–Meyer–Olkin value of 0.879 and Bartlett’s test of sphericity (χ² = 4059.70, p < 0.001). Principal Component Analysis (PCA) with Varimax rotation extracted the four factors, explaining 68% of variance.

The robust methodological framework ensured the constructs effectively captured both cognitive and affective orientations toward AI. Analytical procedures included descriptive statistics, correlation analysis, and regression models to identify relationships between perceptions and demographic or experiential predictors.



Results

Findings indicate ambivalent perceptions of AI among Mashhad’s residents.

Perceived threats were pronounced. Respondents reported anxiety about privacy breaches, unauthorized surveillance, biased or inaccurate decisions in medical contexts, and unemployment risks tied to automation. Average threat scores (≈3.9/7) pointed to cautious attitudes. Concerns extended to misuse for terrorism, systemic bias, and even existential fears of machines surpassing humans.

Trust and comfort showed moderate scores, averaging slightly above 3.9. Some participants expressed confidence in governmental and private institutions deploying AI; however, overall trust was fragile. Familiarity and prior experience correlated positively with trust, while older age was negatively associated. Gender differences showed men reporting marginally higher trust, though not at statistically significant levels.

Applications in urban services were viewed more positively. Citizens acknowledged AI’s potential to reduce costs, improve transportation, optimize resource allocation, and strengthen safety monitoring. Scores in this dimension exceeded 4.2, reflecting greater acceptance. Yet regression analysis revealed a paradox: practical AI experience was negatively associated with perceived usefulness, suggesting unmet expectations among experienced users.

AI’s role in crisis management received the strongest endorsement. Respondents valued AI’s predictive capacity, rapid emergency response, and ability to disseminate accurate crisis information. Items related to early warnings and risk identification scored above 4.0, with nearly half of participants strongly agreeing that AI could substantially enhance emergency response efficiency. Despite limited hands-on experience, citizens were inclined to trust AI in high-stakes scenarios, emphasizing reliance on its objectivity and analytical speed.

Regression results showed AI knowledge and familiarity as key predictors. Knowledge positively influenced trust and comfort, while practical experience simultaneously strengthened trust but tempered optimism about service applications. Demographic factors such as education and income had limited predictive power, demonstrating that perceptions of AI cut across social groups.



Conclusion

The study underscores a complex interplay of optimism and skepticism among Mashhad’s citizens. On the one hand, there is clear acknowledgment of AI’s potential to boost efficiency, lower costs, enhance service delivery, and increase resilience against crises. AI in crisis management was especially valued, as citizens perceived its capacity for life-saving interventions. On the other hand, apprehensions about privacy, job losses, and opaque decision-making remain persistent. The negative relationship between perceived threats and trust emphasizes the importance of transparency, accountability, and public engagement.

Policy implications are significant. Urban managers must prioritize transparent communication, participatory design, and equitable access to AI services. Trust-building requires technical reliability as well as visible ethical and cultural safeguards. Public education to improve technological literacy can reduce uncertainty, while pilot projects showing tangible benefits may bridge the gap between expectations and actual experiences.

Ultimately, Mashhad’s case demonstrates that successful AI deployment depends less on technical sophistication and more on cultivating trust, mitigating risks, and embedding AI in inclusive governance frameworks. AI can become a catalyst for sustainable urban futures only when its use remains ethically grounded, socially responsive, and transparent to citizens.



Funding

There is no funding support.



Authors’ Contribution

Authors contributed equally to the conceptualization and writing of the article.



Conflict of Interest

Authors declared no conflict of interest.



Acknowledgments

We are grateful to all the scientific consultants of this paper.