فصلنامه نوآوری حقوقی

فصلنامه نوآوری حقوقی

هوش مصنوعی و فرایندهای عدالت کیفری: جلوه‌ها و کاربردها

نوع مقاله : مقاله پژوهشی

نویسندگان
1 دانشیار گروه حقوق جزا و جرم‌شناسی، دانشکدۀ حقوق و علوم سیاسی، دانشگاه خوارزمی، تهران، ایران
2 دانشجوی دکتری حقوق جزا و جرم‌شناسی، دانشکدۀ حقوق و علوم سیاسی، دانشگاه خوارزمی، تهران، ایران (نویسندۀ مسئول)
چکیده
نظام‌های عدالت کیفری معاصر در مواجهه با تحولات اجتماعی، گسترش جرائم سازمان‌یافته و پیچیدگی داده‌های جرم با چالش‌هایی مانند ناکامی در کشف جرم، دشواری شناسایی مظنونان، اطالۀ دادرسی و خطاهای انسانی روبه‌رو هستند؛ مشکلاتی که می‌توانند کارایی و مشروعیت عدالت کیفری و حقوق متهمان را تضعیف کنند. ظهور هوش مصنوعی با تحلیل داده‌محور و پردازش سریع اطلاعات، استفاده از فناوری برای بهبود دادرسی عادلانه را مطرح کرده است. مسئلۀ اصلی پژوهش، بررسی تأثیر هوش مصنوعی بر فرایند دادرسی عادلانه و ارزیابی ظرفیت‌ها و مخاطرات آن در کشف جرم، تحقیق و محاکمه است. پرسش اصلی این است که هوش مصنوعی چگونه می‌تواند دادرسی عادلانه را ارتقا دهد و چه تهدیدهایی برای حقوق دفاعی ایجاد می‌کند. فرضیۀ پژوهش بر آن است که استفادۀ ضابطه‌مند از هوش مصنوعی، در چارچوب اصول دادرسی عادلانه، می‌تواند نارسایی‌های موجود را کاهش دهد، مشروط بر آنکه جایگزین تصمیم‌گیری انسانی نشود. پژوهش از نظر ماهیت، توصیفی ـ تحلیلی ثانوی و کاربردی است و داده‌های آن از طریق مطالعۀ کتابخانه‌ای، بررسی متون حقوقی و جرم‌شناختی و تحلیل قوانین و اسناد مرتبط گردآوری شده است. یافته‌ها نشان می‌دهد هوش مصنوعی در کشف جرم می‌تواند با تحلیل داده‌های رفتاری، مکانی و زمانی به شناسایی الگوهای مجرمانه کمک کند؛ در تحقیق نیز داده‌های دیجیتال، ارتباطات الکترونیکی و تصاویر را سریع پردازش کند و خطاهای انسانی را کاهش دهد. در محاکمه، فناوری می‌تواند با تحلیل سوابق و آرای مشابه، نقش دستیار قضائی را ایفا کند و افزایش هماهنگی میان نهادهای قضائی و اجرایی را ممکن سازد. با این ‌حال، جعبۀ سیاه الگوریتمی، سوگیری داده‌ها، تبعیض، دست‌کاری سامانه‌ها و تضعیف شفافیت و حقوق دفاعی مهم‌ترین مخاطرات آن است. نتیجه آنکه هوش مصنوعی باید به‌عنوان ابزار حمایتی و تحلیلی، تحت نظارت انسانی، شفافیت الگوریتمی، مسئولیت‌پذیری و کنترل مستمر به‌کار گرفته شود. تدوین مقررات جامع، توسعۀ مدل‌های توضیح‌پذیر و آموزش قضات و متخصصان نیز ضروری است.
کلیدواژه‌ها
موضوعات

عنوان مقاله English

Artificial Intelligence and Criminal Justice Processes: Manifestations and Applications

نویسندگان English

Majid ghourchibeigi 1
Ghulamali Andishwar 2
1 Associate Professor, Department of Criminal Law and Criminology, Faculty of Law and Political Science, Kharazmi University, Tehran, Iran
2 Ph.D. Student in Criminal Law and Criminology, Faculty of Law and Political Science, Kharazmi University, Tehran, Iran (Corresponding Author)
چکیده English

1. Introduction
Contemporary criminal justice systems continue to encounter significant structural and operational deficiencies that affect both their effectiveness and legitimacy. These deficiencies include failures in crime detection and prevention, unjustified deprivation of liberty, insufficient prosecutorial or investigative capacity to substantiate criminal charges, and judicial errors that may culminate in wrongful convictions or acquittals. Such shortcomings not only impair the proper administration of justice but also erode public trust in legal institutions and undermine confidence in the rule of law. Against this background, the emergence of artificial intelligence represents a significant technological development with the potential to transform multiple stages of criminal proceedings.
Artificial intelligence systems, by virtue of their capacity to process extensive datasets, identify patterns, generate probabilistic assessments, and support complex analytical tasks, are increasingly being considered for application in criminal justice. Such applications include predictive policing, facial recognition, digital surveillance, forensic analysis, risk assessment, document review, and decision-support tools for prosecutors and judges. In principle, these tools may enhance efficiency, improve the management of complex evidentiary materials, reduce certain forms of human error, and promote greater consistency in decision-making. In legal systems burdened by case backlogs, limited institutional capacity, or increasingly sophisticated forms of criminality, the application of these technologies may be particularly attractive.
Nevertheless, the incorporation of artificial intelligence into criminal justice gives rise to significant legal, ethical, and procedural concerns. Criminal procedure is not merely an administrative framework for the efficient processing of cases; rather, it constitutes a rights-oriented system designed to preserve human dignity, ensure due process, safeguard equality before the law, and secure the guarantees of a fair trial. Accordingly, the use of artificial intelligence in this field raises fundamental questions concerning the appropriate limits of technological intervention, the legitimacy of relying on algorithmic systems for decisions traditionally entrusted to human decision-makers, and the compatibility of automated tools with the core principles of procedural justice. Concerns relating to algorithmic bias, discriminatory outcomes, the opacity of automated reasoning, limited explainability, intrusions into privacy, and diminished accountability are particularly acute in criminal matters, where decisions may directly affect individual liberty, reputation, and other fundamental rights.
The present study examines whether, and to what extent, artificial intelligence affects the fairness of criminal proceedings and how the benefits and risks associated with its application should be assessed within the framework of criminal procedure. The principal hypothesis is that, although artificial intelligence may enhance efficiency, analytical capacity, consistency, and relative accuracy in criminal justice, its use may also jeopardize fair-trial guarantees unless subject to robust standards of transparency, explainability, accountability, non-discrimination, and meaningful human oversight.
2. Methodology
In terms of its purpose, this study is applied and adopts a descriptive-analytical approach. The legal dimension of the study is examined through a doctrinal methodology grounded in an analysis of the principles of criminal procedure, legal scholarship, and relevant regulatory and comparative approaches to the use of artificial intelligence in criminal justice. The study also draws on interdisciplinary literature at the intersection of law, technology, ethics, and governance to situate artificial intelligence within broader debates concerning legitimacy, responsibility, and the protection of rights in judicial processes.
The study employs qualitative analysis and inferential interpretation of legal materials and scholarly arguments. For analytical clarity, the role of artificial intelligence in criminal procedure is examined in four principal domains: crime detection and prevention; criminal investigation and evidentiary analysis; adjudication and judicial decision support; and the broader advantages, limitations, and risks associated with the integration of artificial intelligence into criminal proceedings. This framework enables an assessment of artificial intelligence not merely as a technical instrument but also as a normative and institutional phenomenon with significant implications for procedural fairness, judicial legitimacy, and the protection of fundamental rights.
3. Findings
The findings indicate that artificial intelligence has considerable potential to improve the functioning of criminal justice systems in several respects. In the field of crime detection, artificial intelligence may strengthen the capacity of law enforcement authorities to identify suspicious patterns, process large volumes of surveillance and digital data, and detect correlations that may not be readily discernible to human investigators. During the investigative phase, artificial intelligence may facilitate the organization, classification, and analysis of extensive evidentiary materials, including electronic communications, financial records, biometric data, and forensic evidence. These functions may contribute to faster processing, improved evidentiary assessment, and more informed prosecutorial decision-making.
In adjudication, artificial intelligence may also serve as a supplementary decision-support tool, assisting judges in reviewing voluminous case files, identifying relevant legal authorities, structuring legal issues, and summarizing evidentiary records. Such applications may enhance procedural consistency and reduce certain risks associated with cognitive overload, oversights, or inconsistencies in human decision-making. In jurisdictions facing heavy caseloads or institutional inefficiencies, artificial intelligence may therefore offer significant practical benefits.
Nevertheless, the findings also demonstrate that these advantages are accompanied by substantial risks. One of the most significant concerns is the potential of artificial intelligence systems to reproduce, reinforce, or amplify existing social and institutional biases. Where algorithmic models are trained on historically biased or incomplete datasets, their outputs may disproportionately disadvantage particular groups on the basis of race, ethnicity, socioeconomic status, gender, or geographic location. In criminal proceedings, such distortions may undermine equality before the law, weaken the presumption of innocence, and compromise judicial impartiality.
A further challenge concerns algorithmic opacity, commonly characterized as the “black box” problem. In many instances, neither defendants nor legal professionals are able to understand how a particular artificial intelligence system generates a recommendation, classification, or risk assessment. This lack of transparency poses serious obstacles to challenging decisions informed by artificial intelligence, evaluating the reliability of evidence, and ensuring effective judicial scrutiny. The rights of the defense may be materially weakened where an accused person is unable to challenge the reasoning underlying a technological output that influences arrest, detention, the bringing of charges, adjudication, or sentencing.
The findings also reveal concerns relating to privacy, data protection, cybersecurity, and institutional accountability. Artificial intelligence systems used in criminal justice may be vulnerable to technical errors, manipulation, or misuse, while responsibility for harmful outcomes may become diffused among software developers, public authorities, investigators, prosecutors, and judges. Such diffusion is particularly problematic in criminal justice, where the exercise of coercive state power must remain identifiable, legally attributable, and subject to review. Accordingly, the use of artificial intelligence cannot be justified solely on grounds of efficiency or innovation; it must also satisfy the normative requirements of procedural fairness and legal accountability.
4. Conclusion and Recommendations
Conclusion
This study concludes that artificial intelligence should be regarded as a supportive and complementary instrument in criminal justice, rather than as a substitute for human judgment and legal responsibility. Although artificial intelligence may enhance administrative efficiency, analytical capacity, and relative consistency, it cannot replace the interpretive, evaluative, and moral reasoning that is central to criminal adjudication. Decisions concerning guilt, innocence, liberty, and punishment must remain under the effective control of human decision-makers who are legally and ethically accountable for their determinations.
Accordingly, the legitimate use of artificial intelligence in criminal procedure depends on strict adherence to fundamental legal principles, particularly fairness, transparency, explainability, accountability, proportionality, non-discrimination, and effective human oversight. Artificial intelligence should be deployed only in contexts in which its outputs can be meaningfully reviewed, understood, challenged, and corrected within the legal process. The use of opaque or insufficiently auditable systems in core judicial functions should therefore be approached with the utmost caution and should be avoided altogether in certain contexts.
5. Recommendations
On this basis, the study recommends the adoption of clear and comprehensive legal and regulatory frameworks governing the design, deployment, and use of artificial intelligence in criminal justice; the development of explainable, auditable, and rights-sensitive artificial intelligence systems to reduce algorithmic opacity and facilitate meaningful review; the provision of specialized training for judges, prosecutors, lawyers, investigators, and other legal practitioners to enable the critical assessment and proper interpretation of outputs generated by artificial intelligence; the establishment of mechanisms for continuous oversight, independent evaluation, and human rights impact assessment to prevent discriminatory, arbitrary, or unjust outcomes; and the preservation of the principle that artificial intelligence must remain subordinate to human judgment and must not function as an autonomous determinant of criminal liability or punishment.
Accordingly, the future role of artificial intelligence in criminal justice should be shaped not solely by technological optimism, but by a principled balance between innovation and the protection of procedural justice. Only within such a framework may artificial intelligence contribute constructively to criminal proceedings without undermining the essential guarantees of a fair trial.


کلیدواژه‌ها English

Artificial intelligence
Criminal justice
Intelligent judicial system
Technology in justice
دوره 2، شماره 3
بهار 1405
صفحه 107-134

  • تاریخ دریافت 30 آذر 1404
  • تاریخ بازنگری 10 بهمن 1404
  • تاریخ پذیرش 10 اسفند 1404
  • تاریخ اولین انتشار 01 فروردین 1405
  • تاریخ انتشار 01 فروردین 1405