Legal Research Quarterly

Legal Research Quarterly

Capacities and Challenges of AI-Based Video Surveillance Systems in Situational Crime Prevention; Under the European Union Artificial Intelligence Act of 2024

Document Type : Original Article

Authors
1 Associate Professor, Faculty of Law, Shahid Beheshti University, Tehran, Iran.
2 Associate Professor, Department of Philosophy of Science, Sharif University of Technology, Tehran, Iran.
3 PhD Candidate, Faculty of Law, Shahid Beheshti University, Tehran, Iran Corresponding Author Email: s_shafiee@sbu.ac.ir
10.48308/jlr.2026.243090.3027
Abstract
Abstract
Introduction
Recent developments in artificial intelligence (AI) have provided new instruments for strengthening crime prevention. Among them, AI-based video surveillance systems—a combination of CCTV cameras and video analytics algorithms—play a prominent role in governing public spaces and advancing the goals of situational crime prevention. By relying on capabilities such as facial recognition, person and object tracking, abandoned-object detection, suspicious-behavior detection, emotion recognition, and crowd anomaly detection, these systems can reduce opportunities for a wide range of crimes, from theft and vandalism to violent and terrorist offenses. At the same time, these capacities may generate serious risks for fundamental rights and freedoms, privacy, the rule of law, and the presumption of innocence, potentially fostering mass surveillance and diminishing public trust. Accordingly, this article pursues three objectives. First, it offers an innovative account of the preventive capacities of AI-based video surveillance systems across four categories: scene-oriented monitoring, identity-oriented monitoring, individual-oriented monitoring, and crowd-oriented monitoring. This fourfold classification is significant because, although these capacities all operate within the broader framework of video surveillance, they differ in terms of the subject matter of processing, the degree of interference with individual rights, and their preventive functions; therefore, they cannot be subjected to a uniform legal response. Second, it analyzes the corresponding socio-legal challenges. Finally, it evaluates the legal status of each capacity in light of the risk-based approach adopted by the European Union Artificial Intelligence Act (EU AI Act) of 2024.
Methods
This study is descriptive–analytical and adopts an interdisciplinary approach, focusing simultaneously on the technical foundations of AI-based video surveillance systems and the principles of crime prevention. First, the preventive capacities are classified based on the purpose of processing—ranging from situational and pattern monitoring to identity recognition and crowd analysis. Second, the challenges are identified and examined in five main areas: 1) the rule of law, 2) the principle of proportionality, 3) fundamental freedoms, 4) privacy, 5) institutional accountability, including the necessity of human oversight. In the final step, these capacities and challenges are mapped onto the EU AI Act’s risk-based logic to clarify which uses—given the impacts on fundamental rights—fall under prohibitions, strict restrictions, or high-risk system obligations, and which governance mechanisms are required to control their risks. In this analysis, in addition to the technical function of each capacity, its relationship with the objectives of situational crime prevention and its reliance on personal, biometric, or behavioral data are also taken into consideration. Furthermore, in determining their legal status, a distinction is drawn between capacities that are explicitly addressed by the EU AI Act and those that fall only indirectly within the Act’s general provisions.
Results and Discussion
The findings indicate that the development of the technical capacities of AI-based video surveillance systems and the real-time processing of data enable these systems to transform ongoing data into alerts and preventive interventions in real time, thereby thereby narrowing the gap between observation, analysis, and preventive action. These capacities can reinforce the core logic of situational crime prevention—namely, increasing risk for potential offenders, increasing effort, reducing rewards, and reducing provocations. However, their legitimacy and effectiveness depend on the quality of regulation and institutional design. In scene-oriented monitoring, person/object tracking and abandoned-object detection enable early warnings regarding intrusions into restricted areas, suspicious stops, or hazardous items. Yet, false positives and false negatives, together with technical constraints (e.g., occlusion, crowd density, lighting variation), may trigger unnecessary preventive interventions. In identity-oriented monitoring, facial recognition—especially when deployed in real time and at a distance—may enhance deterrence and facilitate targeted control of high-risk individuals. Conversely, because it processes biometric data, it heightens risks of discrimination, misidentification, and privacy infringement; moreover, by weakening the presumption of innocence and the rule of law, it may contribute to transforming citizens into “permanent suspects.” In individual-oriented monitoring, detecting suspicious or violent behavior can support minimal, situational interventions; nevertheless, distinguishing ordinary behavior from crime-relevant conduct remains persistently difficult. Emotion recognition, as a branch of affective computing (emotional AI), is even more contested in terms of scientific foundation and reliability for preventive purposes, and is particularly vulnerable to undermining human dignity. Finally, crowd-oriented monitoring, through crowd anomaly detection without individual identification, may be useful for congestion management and for reducing criminal opportunities in high-density environments; however, in the absence of clear criteria, it may still lead to group labeling and the normalization of mass surveillance. From a legal perspective, the article argues that responding to these challenges requires a combination of strict enforcement of proportionality (suitability, necessity, and proportionality in a strict sense), strengthening the presumption of innocence, protecting fundamental freedoms and privacy, limiting the scope of surveillance, ensuring transparency and prohibiting covert surveillance, institutionalizing accountability and integrating implementing human oversight. Under the EU AI Act (2024), a substantial share of security-related uses involving biometric recognition are treated as high-risk and must therefore comply with enhanced requirements for risk management, data quality, documentation, event logging, human oversight, and mechanisms for accountability and contestation. By contrast, lower-risk uses may be pursued with fewer legal constraints if the above principles are respected. Nevertheless, the Act does not explicitly address certain AI-based capabilities, potentially leaving gaps in legal protection and oversight for individuals’ rights and freedoms.
Conclusion
In conclusion, AI-based video surveillance systems can be an important tool for situational crime prevention, but only when deployed within a clear legal framework and strict institutional controls designed to uphold the rule of law, the presumption of innocence, proportionality, privacy protection, and human oversight. Otherwise, the same preventive capacities may become a threat to fundamental freedoms and human dignity, thereby undermining the legitimacy and effectiveness of crime prevention governance.

Highlights

Highlights

·         AI-based video surveillance systems can strengthen the capacity of situational crime prevention by enabling early intervention.

·         The preventive use of these systems requires adherence to the rule of law, proportionality, fundamental rights and freedoms, privacy, transparency, accountability, and human oversight.

·         The European Union Artificial Intelligence Act of 2024 imposes stringent restrictions on certain applications of AI-based video surveillance systems, while several other capabilities remain in need of supplementary rules and safeguards.

Keywords
Subjects

  1. References

    Books

    1.       Ansari, Bagher, Data Rights and Artificial Intelligence: Concepts and Challenges, Tehran: Sahami Enteshar Company, First Edition, 2021. (In Persian)

    1. Kant, Immanuel, Critique of Practical Reason, Translated by: InshaAllah Rahmati, Tehran: Nour al-Thaqalain, 2006. (In Persian)
    2. Niazpour, Amir Hassan, Crime Prevention, Tehran, Dadgostar, Second Edition, 2023. (In Persian)
    3. Pradel, Jean, History of Penal Doctrines, Translated by: Ali Hossein Najafi Abrand-Abadi, Tehran: Samt, Fourth Edition, 2009. (In Persian)

    Articles

    5.      Akpulat, M., and M. Ekinci, "Anomaly Detection in Crowd Scenes via Cross Trajectories", Applied Intelligence, Vol 55, Issue 525, 2025, pp.1-16. https://doi.org/10.1007/s10489-025-06338-z /

    6.      Ali, M. M., "Real-Time Video Anomaly Detection for Smart Surveillance", IET Image Processing, Vol 17, Issue 5, 2023, pp.1375-1388.https://doi.org/10.1049/ipr2.12720/

    7.      Chen, K., X. Yang, C. Fan, W. Zhang, and Y. Ding, "Semantic-Rich Facial Emotional Expression Recognition", IEEE Transactions on Affective Computing, Vol 13, Issue 4, 2022, pp.1906-1916. https://doi.org/10.1109/TAFFC.2022.3201290/

    8.      Choung, H., P. David, and T-W. Ling. "Acceptance of AI-Powered Facial Recognition Technology in Surveillance Scenarios:Role of Trust, Security, and Privacy Perceptions", Technology in Society, Vol 79, 2024, pp.1-11.https://doi.org/10.1016/j.techsoc.2024.102721/

    9.      Clarke, R. V., "Situational Crime Prevention: Its Theoretical Basis and Practical Scope", Crime and Justice: A Review of Research, University of Chicago Press, Vol 4, 1983, pp.225–256. https://doi.org/10.1086/449090

    10.  Ebers, M., V. R. S. Hoch, F. Rosenkranz, H. Ruschemeier, and B. Steinrötter. "The European Commission’s Proposal for an Artificial Intelligence Act—A Critical Assessment by Members of the Robotics and AI Law Society(RAILS)", J — Multidisciplinary Scientific Journal, Vol 4, Issue 4, 2021, pp.589-603. https://doi.org/10.3390/j4040043/

    11.  Esfahani, Babak and Saber Niavarani, “The Principle of Proportionality; A Tool to Guarantee and Balance the Fundamental Rights and Freedoms and Public Interests”, Legal Research Quarterly, Volume 27, Issue 4, 2024, pp.207-224. (In Persian) https://doi.org/10.48308/jlr.2022.220751.1852/

    12.  Feng, W., F. Wang, R. Han, Y. Gan, Z. Qian, J. Hou, and S. Wang, "Unveiling the Power of Self-Supervision for Multi-View Multi-Human Association and Tracking", IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol 47, Issue 1, 2025, pp.351-368. https://doi.org/10.1109/TPAMI.2024.3463966/ 

    13.  Feretzakis, G., K. Papaspyridis, A. Gkoulalas-Divanis, and V. S. Verykios, "Privacy-Preserving Techniques in Generative AI and Large Language Models: A Narrative Review", Information, Vol 15, Issue 11, 2024, pp.1-25. https://doi.org/10.3390/info15110697/

    14.  Fontes, C., E. Hohma, C. C. Corrigan, and C. Lütge, "AI-Powered Public Surveillance Systems: Why We (Might) Need Them and How We Want Them", Technology in Society, Vol 71, 2022, pp.1-12. https://doi.org/10.1016/j.techsoc.2022.102137 /

    15.  Ghanad, Fatemeh and Elham Sharif, “Comprehensive Study of Personal Data Protection in Iran's Legal System and European General Data Protection Regulations”, Modern Technologies Law Quarterly, Volumw 2, Issue 4, 2021, pp.1-22. (In Persian) https://doi.org/10.22133/clj.2021.244608.1020/

    16.  Guan, Z., Z. Wang, G. Zhang, L. Li, M. Zhang, Z. Shi, and N. Jiang. "Multi-Object Tracking Review: Retrospective and Emerging Trend", Artificial Intelligence Review, Vol 58, Issue 235, 2025, pp.1-46. https://doi.org/10.1007/s10462-025-11212-y/

    17.  Haley, P. "The Impact of Biometric Surveillance on Reducing Violent Crime: Strategies for Apprehending Criminals While Protecting the Innocent", Sensors, Vol 25, Issue 10, 2025, pp.1-21. https://doi.org/10.3390/s25103160 /

    18.  Huang, J., L. Li, L. Qing, W. Tang, P. Wang, L. Guo, and Y. Peng. "Spatio-Temporal Interactive Reasoning Model for Multi-Group Activity Recognition", Pattern Recognition, Vol 159, 2025, pp.1-14. https://doi.org/10.1016/j.patcog.2024.111104/

    19.  Johnson, T. L., N. N. Johnson, V. Topalli, D. McCurdy, and A. Wallace. "Police Facial Recognition Applications and Violent Crime Control in U.S. Cities", Cities, Vol 155, 2024, pp.1-14. https://doi.org/10.1016/j.cities.2024.105472 /

    20.  Matei, A-D., E. Talavera, and M. Aghae."Crime Scene Classification from Skeletal Trajectory Analysis in Surveillance Settings", Engineering Applications of Artificial Intelligence, Vol 141, 2025, pp.1-14. https://doi.org/10.1016/j.engappai.2024.109800

    21.  Podoletz, L. "We Have to Talk about Emotional AI and Crime", AI & Society, Vol 38, Issue 3, 2023, pp.1067–1082. https://doi.org/10.1007/s00146-022-01435-w/

    22.  Rezaee, Kh, S. Rezakhani, M. R. Khosravi, and M. K. Moghimi. "A Survey on Deep Learning-Based Real-Time Crowd Anomaly Detection for Secure Distributed Video Surveillance", Personal and Ubiquitous Computing, Vol 28, Issue 1, 2024, pp.135–151.https://doi.org/10.1007/s00779-021-01586-5/

    23.  Serna, I., A. Morales, J. Fierrez, and N. Obradovich. "Sensitive Loss: Improving Accuracy and Fairness of Face Representations with Discrimination-aware Deep Learning", Artificial Intelligence, Vol 305, 2022, pp.1-15.https://doi.org/10.1016/j.artint.2022.103682/

    24.  Simmler, M., and G. Canova. "Facial Recognition Technology in Law Enforcement: Regulating Data Analysis of Another Kind", Computer Law & Security Review, Vol 56, 2025, pp.1-10. https://doi.org/10.1016/j.clsr.2024.106092/

    25.  Talaei Khoei, T., and A. Singh. "A Survey of Emotional Artificial Intelligence and Crimes: Detection, Prediction, Challenges and Future Direction", Journal of Computational Social Science, Vol 7, 2024, pp.2359–2402. https://doi.org/10.1007/s42001-024-00313-3/

    26.  Veltmeijer, E., and C. Gerritsen. "Legal and Ethical Implications of AI-based Crowd Analysis: The AI Act and Beyond", AI Ethics, Vol 5, Issue 3, 2025, pp.3173–3183. https://doi.org/10.1007/s43681-024-00644-x/

    27.  Welsh, B. C., and D. P. Farrington. "Surveillance for Crime Prevention in Public Space: Results and Policy Choices in Britain and America", Criminology and Public Policy, Vol 3, Issue 3, 2004, pp.497–526. https://doi.org/10.1111/j.1745-9133.2004.tb00054.x/ 

    28.  Zaidenberg, S., B. Boulay, and F. Brémond. "A Generic Framework for Video Understanding Applied to Group Behavior Recognition", IEEE Ninth International Conference on Advanced Video and Signal-Based Surveillance, 2012, pp.1-7. https://doi.org/10.1109/AVSS.2012.1/

    29.  Zhou, L., and J. Xu. "Enhanced Abandoned Object Detection through Adaptive Dual-Background Modeling and SAO-YOLO Integration", Sensors, Vol 24, Issue 20, 2024, pp.1-20. https://doi.org/10.3390/s24206572

    Reports

    30.  Ratcliffe, Jerry, "Video Surveillance of Public Places (Problem-Oriented Guides for Police. Response Guide Series. Guide No. 4)", U.S.Department of Justice Office of Community Oriented Policing Services, Washington, DC: Center for Problem-Oriented Policing, 2006. https://popcenter.asu.edu/sites/g/files/litvpz3631/files/video_surveillance.pdf

    31.  The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems. Ethically Aligned Design: A Vision for Prioritizing Human Well-being with Autonomous and Intelligent Systems, First Edition. IEEE, 2019.https://standards.ieee.org/content/ieee-standards/en/industry-connections/ec/autonomous-systems.html/

    Website

    32.     "CCTV Camera Market Size, Global Report 2022-2029", Fortune Business Insights, 2025, Accessed 02 December 2025. https://www.fortunebusinessinsights.com/cctv-camera-market-107115/

  • Receive Date 13 January 2026
  • Revise Date 15 February 2026
  • Accept Date 18 March 2026