Debido proceso, sesgo cognitivo e inteligencia artificial en la toma de decisiones judiciales J. Law Epistemic Stud. (2026) 4: e167 https://doi.org/10.5281/zenodo.20059260 ISSN 3091-1575 ORIGINAL ARTICLE Due process, cognitive bias and artificial intelligence in judicial decision-making  Ariadna Moya- Rodríguez ariadna.moya@unic.co.ao Received: 01 February 2026 / Accepted: 21 March 2026 / Published online: 30 April 2026 © The Author(s) 2026 Ariadna Moya- Rodríguez 1,2 Yonayka Licea -Suárez 1,2 Abstract This study analyzes the legal and psychological impli- cations of artiﬁcial intelligence in judicial decision-making, with particular emphasis on its eﬀects on due process guarantees. The research adopts a qualitative, doctrinal, and comparative approach based on a systematic review of academic literature and interna- tional regulatory frameworks related to AI governance and digital justice. From a legal perspective, the ﬁndings identify risks asso- ciated with algorithmic opacity, automated bias, and the weaken- ing of reasoned judicial decisions, potentially aﬀecting equality before the law, judicial impartiality, and the right to defense. From a psychological perspective, the study highlights the inﬂuence of cognitive biases, particularly automation bias and anchoring ef- fects, on judicial actors interacting with AI systems, which may reduce autonomy and critical reasoning. The results also indicate that procedural fairness and public trust in judicial institutions may be undermined when algorithmic processes lack transparency and explainability. In response to these challenges, the study propos- es a hybrid model of adjudication in which artiﬁcial intelligence functions as a decision-support tool under strict human supervision, ensuring the protection of fundamental rights and the preservation of cognitive integrity in judicial decision-making processes. Keywords Artiﬁcial intelligence, due pro- cess, judicial decision making, cognitive bias algorithmic transparency. Resumen El presente estudio analiza las implicaciones jurídicas y psicológicas de la inteligencia artiﬁcial en la toma de decisiones judiciales, con especial énfasis en sus efectos sobre las garantías del debido proceso. La investigación adopta un enfoque cualitati- vo, doctrinal y comparado basado en una revisión sistemática de literatura cientíﬁca y marcos regulatorios internacionales relaciona- dos con la gobernanza de la inteligencia artiﬁcial y la justicia digi- tal. Desde la perspectiva jurídica, los hallazgos identiﬁcan riesgos asociados a la opacidad algorítmica, los sesgos automatizados y el debilitamiento de la motivación de las decisiones judiciales, afec- tando la igualdad ante la ley, la imparcialidad judicial y el derecho a la defensa. Desde la perspectiva psicológica, el estudio destaca la inﬂuencia de sesgos cognitivos, particularmente el sesgo de au- tomatización y el efecto de anclaje, sobre los operadores jurídicos que interactúan con sistemas de inteligencia artiﬁcial, lo que puede reducir la autonomía y el razonamiento crítico. Los resultados tam- bién indican que la imparcialidad procesal y la conﬁanza pública en las instituciones judiciales pueden verse socavadas cuando los procesos algorítmicos carecen de transparencia y explicabilidad. En respuesta a estos desafíos, el estudio propone un modelo hí- brido de adjudicación en el que la inteligencia artiﬁcial funciona como una herramienta de apoyo a la toma de decisiones bajo estric- ta supervisión humana, garantizando la protección de los derechos fundamentales y la preservación de la integridad cognitiva en los procesos de toma de decisiones judiciales. Palabras clave Inteligencia artiﬁcial, debido proce- so, toma de decisiones judiciales, sesgos cognitivos transparencia algorítmica. How to cite Moya-Rodríguez, A., & Licea-Suárez, Y. (2026). Due process, cognitive bias and artiﬁcial intelligence in judicial decision-making. Journal of Law and Epistemic Studies, 4, e167. https://doi.org/10.5281/zenodo.20059260 1 Fundación Iberoamericana Universitaria (FUNIBER), España. 2 Universidade Internacional do Cuanza: Cuito, Bíe, Angola.

J. Law Epistemic Stud. (January - June) 4: e167 Introduction Due process of law constitutes a foundational guarantee of the constitutional rule of law, insofar as it imposes nor- mative constraints upon the exercise of public authority and ensures that any determination aﬀecting individual rights is rendered through fair, transparent, and duly reasoned pro- cedures. Traditionally anchored in principles such as judi- cial impartiality, the right to be heard (audi alteram partem), procedural equality, and the duty to provide reasoned deci- sions, due process has progressively evolved in response to institutional developments and, more recently, to the rapid advancement of digital technologies within legal systems. In this regard, the integration of artiﬁcial intelligence AI into judicial administration represents a paradigmatic shift in the manner in which adjudicative functions are performed. Contemporary AI systems particularly those grounded in machine learning, natural language processing, and pre- dictive analytics are increasingly deployed to assist, and in certain instances partially automate, decision making pro- cesses. These applications range from case management and legal research to risk assessment and outcome prediction, reﬂecting the broader expansion of legal technology aimed at enhancing eﬃciency and reducing judicial backlog (Siva- satyanarayanareddy, 2024). Notwithstanding these potential beneﬁts, the growing reliance on algorithmic systems raises substantial concerns regarding their compatibility with the procedural safeguards inherent in due process. From a jurisprudential perspective, AI must not be con- ceived merely as a neutral or auxiliary tool, but rather as a socio technical infrastructure capable of shaping legal out- comes and redistributing decision making authority. In this vein, Novelli et al (2026) contends that AI systems should be evaluated within a broader framework of information eth- ics, recognizing their capacity to inﬂuence the protection of fundamental rights and the architecture of legal accountabil- ity. Kumar et al (2025) underscores the risks associated with opaque algorithmic systems commonly referred to as black boxes which hinder meaningful oversight and undermine the principles of transparency and accountability essential to the rule of law. A central doctrinal concern arises from the tension be- tween algorithmic opacity and the duty to provide reasoned judgments. Due process requires that judicial decisions be supported by intelligible, reviewable, and legally grounded reasoning. However, many AI systems particularly those based on complex statistical models operate through pro- cesses that are neither readily interpretable nor susceptible to eﬀective scrutiny. As Fisher (2025) has demonstrated, such opacity poses signiﬁcant legal challenges, as it impairs the ability of aﬀected parties to contest decisions and un- dermines the guarantees of procedural fairness and eﬀective judicial protection. Moreover, the emergence of algorithmic bias constitutes a critical threat to the principle of equality before the law. Empirical research, notably the study conducted by Neil, and Zanger-Tishler, (2025), concerning the COMPAS risk assessment tool, has revealed that algorithmic systems may reproduce and even amplify existing structural inequalities, particularly in criminal justice contexts. These ﬁndings call into question the neutrality of automated decision making and highlight the necessity of subjecting such systems to rig- orous legal and ethical scrutiny. From a psychological perspective, the incorporation of artiﬁcial intelligence into judicial decision making intro- duces additional layers of complexity that extend beyond normative concerns. Research in cognitive psychology has demonstrated that human decision making is inherently in- ﬂuenced by heuristics and biases, including anchoring ef- fects, availability heuristics, and overreliance on automated systems, commonly referred to as automation bias (Rosbach 2026; Romeo & Conti, 2026). In judicial contexts, these cog- nitive tendencies may aﬀect how judges interpret and rely upon algorithmic recommendations, potentially diminishing critical evaluation and independent reasoning. Furthermore, the perception of procedural fairness among litigants is closely linked to psychological factors such as transparency, explainability, and the opportunity to challenge decisions. When algorithmic systems operate without suﬃcient clari- ty, they may not only compromise legal guarantees but also erode trust in judicial institutions, thereby aﬀecting the per- ceived legitimacy of the decision making process. From the standpoint of legal theory, these developments prompt a reassessment of the nature of adjudication itself. While traditional judicial reasoning entails interpretative judgment, normative balancing, and contextual analysis, AI-driven systems rely predominantly on probabilistic infer- ence and pattern recognition. This shift raises the concern that legal decision making may increasingly be guided by predictive outputs rather than principled legal reasoning, thereby eroding the normative foundations of adjudication (Khokhar et al., 2025) At the international level, regulatory and policy frame- works have begun to address these challenges. Instruments such as those developed by Ayibam (2025) emphasize the necessity of ensuring that AI deployment within judicial systems adheres to core principles including transparency, accountability, non discrimination, and meaningful human oversight. These developments reﬂect an emerging consen- sus that technological innovation must operate within the boundaries imposed by fundamental rights and the rule of law. In the Latin American context, the adoption of AI in judi- cial systems occurs against a backdrop of structural inequal-

J. Law Epistemic Stud. (January - June) 4: e167 ities, institutional constraints, and uneven technological development. While AI presents opportunities to enhance ac- cess to justice and administrative eﬃciency, its unregulated implementation risks exacerbating existing disparities and undermining procedural guarantees (Blount, 2026). Conse- quently, any normative assessment must be attentive to the speciﬁc socio legal conditions in which such technologies are deployed. Despite the growing body of literature on artiﬁcial intel- ligence in judicial systems, existing studies have predom- inantly focused on technological eﬃciency, algorithmic accountability, and regulatory governance. Comparatively limited attention has been devoted to the interdisciplinary in- teraction between procedural due process and the cognitive dynamics aﬀecting judicial actors in AI-assisted environ- ments. In particular, there remains a signiﬁcant gap regard- ing how automation bias, anchoring eﬀects, and algorithmic opacity jointly inﬂuence both the normative legitimacy and psychological integrity of judicial decision-making. This study seeks to address this gap by integrating legal and cog- nitive perspectives into a uniﬁed analytical framework for assessing AI-assisted adjudication. Against this backdrop, the present study is guided by the following research question to what extent does the imple- mentation of artiﬁcial intelligence in judicial administration aﬀect due process guarantees and under what normative conditions can such technologies be reconciled with a rights based model of adjudication. The primary objective is to un- dertake a critical analysis of this relationship, identifying the principal risks associated with AI deployment and proposing regulatory criteria aimed at safeguarding fundamental pro- cedural rights. The working hypothesis posits that, although AI systems may contribute to increased eﬃciency in judicial processes, their deployment in the absence of adequate safeguards par- ticularly with respect to transparency, explainability, and hu- man oversight poses a signiﬁcant risk to essential elements of due process, including the duty to provide reasoned deci- sions, judicial impartiality, and the right to defense. Accord- ingly, this study advocates for the development of a hybrid model of adjudication, wherein AI functions as a decision support tool under strict human supervision, ensuring full compliance with constitutional and procedural guarantees. Methodology This study adopts a qualitative, doctrinal, and comparative research design aimed at critically examining the implica- tions of artiﬁcial intelligence AI for due process guarantees within judicial systems. The research is grounded in a legal analytical approach, supported by an interdisciplinary pers- pective that integrates insights from legal theory, technology studies, public governance, and cognitive psychology. Such an approach is particularly appropriate for addressing com- plex normative and behavioral questions arising from the interaction between emerging technologies and fundamental procedural rights. The methodological framework is primarily based on doc- trinal legal analysis, which involves the systematic exami- nation of legal principles, norms, and jurisprudential doctri- nes related to due process. Through this method, the study identiﬁes and interprets core procedural guarantees such as the right to a fair hearing, the duty to provide reasoned decisions, judicial impartiality, and equality before the law, and evaluates how these guarantees may be aﬀected by the incorporation of AI into judicial decision making processes. This analytical approach allows for a rigorous assessment of the compatibility between algorithmic systems and constitu- tional standards. In addition, the research incorporates a structured review of academic literature published between 2020 and 2025, with particular emphasis on peer reviewed journals indexed in Scopus and Web of Science. The review focuses on key the- mes such as algorithmic accountability, explainability, bias in automated decision making, and the legal implications of AI in judicial contexts. Furthermore, it integrates contribu- tions from cognitive psychology and behavioral sciences to examine how heuristics, cognitive biases, and human-machi- ne interaction inﬂuence judicial reasoning in environments assisted by artiﬁcial intelligence. Relevant sources include leading journals such as Artiﬁcial Intelligence and Law and Computer Law and Security Review, as well as foundational works in decision making theory. Furthermore, a comparative legal perspective is employed to examine regulatory and policy frameworks governing the use of AI in judicial systems across diﬀerent jurisdictions. This includes the analysis of international guidelines such as those developed by UNESCO, as well as emerging regula- tory approaches in Europe and Latin America. The compara- tive dimension enables the identiﬁcation of common princi- ples, divergences, and best practices in the legal governance of AI, thereby enriching the normative and interdisciplinary evaluation undertaken in this study. The analysis is structured around a normative framework based on fundamental due process guarantees, which serve as evaluative criteria for assessing the legitimacy of AI-as- sisted decision-making. These include the right to be heard audi alteram partem, the requirement of reasoned judicial decisions, the principle of judicial impartiality, and equality before the law. In parallel, the study incorporates a psycho- logical analytical dimension focused on cognitive bias, auto- mation bias, and perception of procedural fairness, allowing for a more comprehensive assessment of how judicial actors interact with algorithmic systems. Finally, it is important to acknowledge certain limitations

J. Law Epistemic Stud. (January - June) 4: e167 inherent in this research. As a qualitative and doctrinal study, it does not incorporate empirical or quantitative data, which may limit the generalizability of its ﬁndings. Moreover, given the rapid evolution of AI technologies and regulatory frameworks, some recent developments may not yet be fully reﬂected in the available literature. Notwithstanding these limitations, the study provides a robust and coherent analytical basis for understanding both the normative and psychological challenges posed by AI to due process in contemporary legal systems. The literature review was conducted between January and March 2026 using the Scopus, Web of Scien- ce, SSRN, and Google Scholar databases. The search process employed combinations of keywords such as “artiﬁcial intelligence”, “due process”, “judicial deci- sion-making”, “algorithmic bias”, “automation bias”, “procedural fairness”, and “explainable AI”. Inclusion criteria comprised peer-reviewed articles, legal reports, and academic book chapters published between 2020 and 2026 in English or Spanish. Sour- ces were selected based on academic relevance, cita- tion impact, methodological rigor, and direct relation to AI-assisted adjudication and procedural guarantees. Duplicate records and studies unrelated to judicial deci- sion-making contexts were excluded. Results and discussion The ﬁndings of this study reveal that the incorpora- tion of artiﬁcial intelligence AI into judicial systems produces a dual eﬀect on the functioning of due pro- cess. On the one hand, it enhances procedural eﬃcien- cy and decision making capacity, while on the other it introduces structural and cognitive risks that directly aﬀect core procedural guarantees. This duality reﬂects a fundamental tension between technological optimiza- tion, normative safeguards, and human decision making processes. As shown in Table 1, the risks associated with AI-as- sisted adjudication extend beyond purely technological concerns and involve both normative and cognitive di- mensions. The ﬁndings indicate that algorithmic opaci- ty and automation bias directly aﬀect judicial autonomy, procedural fairness, and the legitimacy of legal reason- ing. Consequently, the governance of AI within judicial systems requires interdisciplinary safeguards capable of protecting both constitutional guarantees and human cognitive integrity. From an operational perspective, AI systems con- tribute signiﬁcantly to the modernization of judicial administration. The automation of repetitive tasks, the capacity to process large volumes of legal data, and the use of predictive analytics enable courts to reduce case backlogs and improve procedural timelines. These ad- vantages are particularly relevant in jurisdictions char- acterized by structural ineﬃciencies and limited institu- tional capacity. In this sense, AI may be understood as a tool that promotes access to justice by facilitating faster and more consistent decision making processes. However, these beneﬁts are accompanied by sub- stantial risks that challenge the normative foundations of due process. One of the most critical issues identiﬁed is algorithmic opacity. Many AI systems operate through complex models that lack transparency, making it diﬃcult for judges, litigants, and oversight bodies to understand the reasoning underlying automated or assisted decisions. This opacity undermines the duty to provide reasoned judgments, which is a cornerstone of procedural fairness and judicial accountability. As emphasized by Mugamba (2025), the pro- liferation of opaque decision making systems threatens the possibility of eﬀective legal scrutiny and democratic control. Closely related to this issue is the presence of algorithmic bias, which poses a direct risk to the principle of equality before the law. Empirical evidence indicates that AI systems trained on historical data may reproduce and amplify exist- ing social inequalities. The well documented case analyzed Table 1. Interdisciplinary Risks of AI-Assisted Judicial Decision-Making Dimension Legal Implication Psychological Eﬀect Due Process Risk Algorithmic opacity Lack of explainability and limited judicial scrutiny Distrust and reduced conﬁdence in judicial outcomes Weakening of defense rights and procedural transparency Automation bias Excessive reliance on algorithmic recommendations Reduced critical reasoning and cognitive dependency Risk of judicial partiality and diminished autonomy Predictive analytics Statistical dependence in legal reasoning Anchoring eﬀect and constrained deliberation Erosion of reasoned judicial decisions Algorithmic discrimination Unequal treatment and potential discriminatory outcomes Perceived unfairness and institutional distrust Violation of equality before the law Lack of human oversight Accountability gaps and insuﬃcient reviewability Reduced judicial autonomy Weakening of judicial legitimacy and democratic accountability

J. Law Epistemic Stud. (January - June) 4: e167 Yeung, and Hartmann (2025) demonstrates how risk assess- ment algorithms in criminal justice contexts can generate discriminatory outcomes, particularly along racial and socio economic lines. This ﬁnding is consistent with the arguments advanced by (Selbst, 2025), who contend that data driven systems may produce disparate impacts even in the absence of explicit discriminatory intent. From a psychological perspective, the results suggest that the use of AI in judicial decision making may generate cog- nitive distortions that aﬀect the autonomy of judicial reason- ing. One of the most relevant phenomena is automation bias, which refers to the tendency of individuals to over rely on automated systems, even when such systems are fallible. In judicial contexts, this may lead judges to give disproportion- ate weight to algorithmic recommendations, reducing critical analysis and independent evaluation of the case. In addition, anchoring eﬀects may arise when algorithmic outputs serve as initial reference points that shape subsequent reasoning, thereby limiting the scope of deliberation (Heo, 2025). Another signiﬁcant concern relates to the potential trans- formation of judicial reasoning as a normative practice. Tra- ditional adjudication is grounded in interpretative reasoning, the balancing of principles, and contextual analysis. By con- trast, AI systems rely on probabilistic inference and pattern recognition, which may shift the logic of decision making from justiﬁcation to prediction. This transformation raises the risk that judicial decisions become increasingly depen- dent on statistical outputs rather than legal reasoning, there- by weakening the epistemic and normative foundations of adjudication. Braun (2025), argues, the ethical governance of AI must ensure that technological systems do not displace human judgment in domains where responsibility and ac- countability are essential. Furthermore, the ﬁndings indicate that the use of AI may aﬀect the right to defense and the principle of adversarial proceedings not only from a legal standpoint but also from a psychological perspective. When decisions are inﬂuenced by algorithmic outputs that are not transparent or explainable, litigants may be unable to understand or challenge the basis of those decisions. This limitation aﬀects not only procedur- al guarantees but also the perception of fairness, which is a central element in the legitimacy of judicial systems. Re- search in procedural justice has shown that individuals are more likely to accept decisions when they perceive the pro- cess as fair, transparent, and participatory. From a comparative perspective, the analysis of interna- tional frameworks reveals an emerging consensus on the need to regulate AI in accordance with fundamental rights principles. Instruments such as those developed by UNES- CO emphasize the importance of transparency, accountabil- ity, non discrimination, and meaningful human oversight in the deployment of AI systems. These principles are increas- ingly reﬂected in regulatory initiatives, particularly within the European context, where eﬀorts are underway to estab- lish binding standards for trustworthy AI. In light of these ﬁndings, it becomes evident that the in- tegration of AI into judicial systems cannot be approached solely as a matter of technological innovation. Rather, it requires a comprehensive normative and interdisciplinary framework that ensures the compatibility of algorithmic tools with due process guarantees and human cognitive integrity. The concept of a hybrid model of adjudication emerges as a viable solution, whereby AI is employed as a decision support mechanism under strict human supervision. Such a model preserves the beneﬁts of technological eﬃ- ciency while maintaining the central role of judicial reason- ing, accountability, and cognitive independence. Ultimately, the results underscore that the legitimacy of AI in judicial contexts depends not only on its technical per- formance but also on its alignment with constitutional prin- ciples and its impact on human decision making processes. The challenge lies in designing regulatory and institutional mechanisms that enable the responsible use of AI without compromising the fundamental rights and psychological conditions that underpin the rule of law. Toward a Hybrid Model of AI-Assisted Adju- dication The ﬁndings of this study suggest that neither fully au- tomated adjudication nor exclusively traditional judicial decision-making provides an adequate response to the con- temporary challenges posed by artiﬁcial intelligence in le- gal systems. In this context, a hybrid model of AI-assisted adjudication emerges as a normative and interdisciplinary framework capable of reconciling technological eﬃciency with due process guarantees, judicial accountability, and cognitive integrity. Unlike purely technological approaches, the proposed model recognizes that the legitimacy of judicial deci- sion-making depends not only on procedural safeguards but also on the preservation of independent human reasoning and critical evaluation in AI-assisted environments. A. Legal safeguards dimension From a legal perspective, the hybrid model requires that all AI-assisted judicial decisions remain subject to meaningful human review, procedural transparency, and constitutional accountability. Algorithmic systems must therefore oper- ate within clear legal boundaries that preserve fundamental guarantees such as the right to defense, equality before the law, judicial impartiality, and the duty to provide reasoned decisions.

J. Law Epistemic Stud. (January - June) 4: e167 In this regard, explainability and reviewability constitute essential conditions for the legitimate use of artiﬁcial intel- ligence within judicial systems. Decisions inﬂuenced by al- gorithmic systems must remain understandable, contestable, and subject to judicial scrutiny in accordance with rule-of- law principles. B. Cognitive safeguards dimension Beyond normative guarantees, the proposed model incor- porates a cognitive safeguard dimension aimed at mitigating the psychological risks associated with AI-assisted adjudi- cation. Existing research demonstrates that automation bias, anchoring eﬀects, and excessive reliance on algorithmic rec- ommendations may weaken independent judicial reasoning and reduce critical analysis during decision-making process- es. Consequently, the model emphasizes the need to preserve cognitive autonomy among judicial actors through mecha- nisms such as bias-awareness training, critical reasoning re- inforcement, and institutional protocols designed to prevent overdependence on automated systems. Within this frame- work, artiﬁcial intelligence functions as an advisory tool rather than a substitute for human judgment. C. Technological safeguards dimension From a technological standpoint, the hybrid model pri- oritizes the implementation of interpretable, auditable, and transparent AI systems capable of providing traceable de- cision pathways. Algorithmic processes should therefore be designed according to principles of explainable artiﬁcial intelligence (XAI), enabling judges, litigants, and oversight bodies to understand the rationale underlying AI-assisted outputs. The incorporation of auditability and algorithmic trace- ability mechanisms is particularly important for ensuring procedural legitimacy and preventing arbitrary or discrim- inatory outcomes. D. Institutional safeguards dimension At the institutional level, the framework requires the es- tablishment of regulatory oversight mechanisms, judicial ethics protocols, and specialized governance structures ca- pable of supervising the responsible use of AI within judicial systems. This includes the development of ethical review committees, transparency standards, and interdisciplinary training programs aimed at strengthening both legal and technological competencies among judicial actors. In addition, several international regulatory instruments have established normative standards aimed at ensuring the responsible and rights-based deployment of artiﬁcial intelli- gence within public institutions and judicial systems. Among the most signiﬁcant developments is the EU AI Act adopted by the European Union, which introduces a risk-based reg- ulatory framework for artiﬁcial intelligence and establishes speciﬁc obligations concerning transparency, accountability, human oversight, and the protection of fundamental rights in high-risk AI systems. Likewise, the UNESCO Recommendation on the Ethics of Artiﬁcial Intelligence emphasizes the need to guarantee hu- man dignity, fairness, transparency, non-discrimination, and democratic governance in the development and deployment of AI technologies. In the judicial sphere, these principles are complemented by the European Ethical Charter on the Use of Artiﬁcial Intelligence in Judicial Systems, which highlights respect for fundamental rights, non-discrimination, quality and security, transparency, impartiality, and user control as essential safeguards for trustworthy judicial AI systems. These international instruments reinforce the argument that the legitimacy of AI-assisted adjudication depends not only on technological sophistication, but also on the existence of robust institutional, ethical, and constitutional safeguards ca- pable of preserving due process and judicial autonomy. Ultimately, the hybrid model proposed in this study seeks to preserve the central role of human judicial reasoning while simultaneously beneﬁting from the operational advantages oﬀered by artiﬁcial intelligence. Rather than replacing judg- es with automated systems, the framework advocates for a collaborative relationship in which technological innovation remains subordinated to constitutional guarantees, cognitive integrity, and democratic accountability. In this sense, the legitimacy of AI-assisted adjudication depends not on the sophistication of algorithmic systems alone, but on the capacity of legal institutions to ensure that technological eﬃciency does not undermine the fundamental principles of due process and the rule of law. Figure 1 illustrates the interdisciplinary structure of the proposed hybrid model of AI-assisted adjudication, inte- grating legal, cognitive, technological, and institutional safeguards aimed at preserving due process guarantees and judicial autonomy. This study contributes to the emerging literature on AI-assisted adjudication by integrating legal and cognitive dimensions into a uniﬁed due process framework. Unlike predominantly technological approaches, the proposed mod- el emphasizes that judicial legitimacy depends not only on algorithmic transparency but also on the preservation of cog- nitive autonomy and independent legal reasoning.

J. Law Epistemic Stud. (January - June) 4: e167 Conclusions The integration of artiﬁcial intelligence AI into judicial systems represents a profound transformation in the admi- nistration of justice, with signiﬁcant implications for the structure and guarantees of due process. This study has de- monstrated that, while AI oﬀers substantial beneﬁts in terms of eﬃciency, consistency, and optimization of judicial pro- cesses, its incorporation also introduces complex normative and psychological challenges that cannot be overlooked. From a legal perspective, the ﬁndings conﬁrm that the most critical risks arise from algorithmic opacity, bias in automa- ted decision making, and the potential weakening of the duty to provide reasoned judicial decisions. These factors directly aﬀect core elements of due process, particularly judicial im- partiality, equality before the law, and the right to defense. In this sense, the deployment of AI in judicial contexts must be understood not merely as a technological advancement, but as a transformation with deep constitutional implications that requires careful regulatory design. At the same time, the study highlights that the challenges posed by AI extend beyond the normative dimension and signiﬁcantly impact the cognitive processes underlying ju- dicial decision making. The presence of automation bias, anchoring eﬀects, and overreliance on algorithmic recom- mendations may reduce the autonomy and critical capacity of judicial actors. These psychological dynamics introduce a new layer of risk, as they can subtly inﬂuence decisions even when formal legal guarantees appear to be preserved. Consequently, the protection of due process must also consi- der the cognitive conditions under which judicial reasoning is exercised. A central contribution of this research lies in articulating a comprehensive interdisciplinary framework that integra- tes legal and psychological perspectives in the analysis of AI-assisted adjudication. Within this framework, the concept of a hybrid model of adjudication emerges as a viable and necessary approach. This model advocates for the use of AI as a decision support tool under strict human supervision, ensuring that ﬁnal authority remains with judicial actors and that reasoning processes remain transparent, accountable, and subject to review. Furthermore, the study underscores the importance of de- veloping regulatory and institutional safeguards grounded in fundamental rights principles. These safeguards must gua- rantee transparency, explainability, accountability, and non discrimination, while also addressing the psychological di- mensions of trust, legitimacy, and perception of fairness. In this regard, the legitimacy of AI in judicial systems depends not only on its technical performance but also on its ability to preserve both the normative and cognitive integrity of the decision making process. From a broader perspective, the ﬁndings suggest that the future of digital justice will depend on the capacity of legal systems to reconcile technological innovation with the pro- Figure 1. Hybrid Model of AI-Assisted Adjudication. Note. Prepared by the authors based on the interdisciplinary analysis of due process, cognitive bias, and AI-assisted adjudi- cation.

J. Law Epistemic Stud. (January - June) 4: e167 tection of human judgment. The challenge is not simply to incorporate artiﬁcial intelligence into judicial systems, but to govern it in a way that strengthens rather than weakens the foundational principles of the rule of law. Finally, this study acknowledges certain limitations, par- ticularly the absence of empirical data and the rapid evolu- tion of AI technologies and regulatory frameworks. Future research should therefore focus on empirical analyses of ju- dicial behavior in AI-assisted environments, as well as on the development of measurable indicators to assess the impact of algorithmic systems on both due process guarantees and cognitive decision making dynamics. In conclusion, artiﬁcial intelligence has the potential to enhance judicial systems, but its legitimacy ultimately de- pends on its alignment with due process principles and its compatibility with the psychological conditions that sustain fair and reasoned decision making. Only through a carefully designed interdisciplinary framework can AI contribute to a more eﬃcient, equitable, and trustworthy system of justice. Ultimately, the future legitimacy of AI-assisted adjudica- tion will depend not on the replacement of human judges by intelligent systems, but on the capacity of legal institutions to preserve constitutional guarantees, cognitive autonomy, and democratic accountability within increasingly automa- ted judicial environments. Additional limitations include the rapid evolution of AI re- gulatory frameworks, limited access to proprietary judicial algorithms, and the absence of cross-jurisdictional empirical evidence regarding AI-assisted adjudication. References Ayibam, J. N. (2025). Artiﬁcial Intelligence in Public Pro- curement: Legal Frameworks, Ethical Challenges, and Policy Solutions for Transparent and Eﬃcient Gover- nance. Alkebulan: A Journal of West and East African Studies, 5(2), 54-69. http://gnosipublishers.com.ng/in- dex.php/alkebulan/article/view/24?captcha Blount, K. (2026). Toward Equal Access to Justice for All: Harnessing Artiﬁcial Intelligence for the Administra- tion of Criminal Justice. Peace, Justice and Strong Insti- tutions. In I. A. Ibrahim, J. Truby, & P. D. Farah (Eds.), Artiﬁcial Intelligence, Governance and Sustainable De- velopment (pp. 185-215). Routledge. Boonprakong, N. (2025). Quantifying and Understanding the Eﬀects of Cognitive Biases in Human-Computer In- teraction. [Doctoral thesis, University of Melbourne]. Repository of The University of Melbourn. https://hdl. handle.net/11343/361041 Braun, T. (2025). Liability for artiﬁcial intelligence reason- ing technologies – a cognitive autonomy that does not help. Corporate Governance. https://doi.org/10.1108/ CG-09-2024-0471 Council of Europe. (2018). European Ethical Charter on the Use of Artiﬁcial Intelligence in Judicial Sys- tems and their environment. European Commission for the Eﬃciency of Justice (CEPEJ). https://rm.coe. int/ethical-charter-en-for-publication-4-decem- ber-2018/16808f699c Fisher, T. (2025). Within the Sound of Silence: Reassess- ing the Role of Reasoning in Judicial Decision-Mak- ing. Cambridge Law Review, 10(1), 1-24. https:// heinonline.org/HOL/Page?handle=hein.journals/cam- brilv10&div=6&id=&page= Heo, W. (2025). Revisiting Algorithmic Bounded Rational- ity with Potential Solutions: How Recommendation Systems Amplify Cognitive Biases in Consumer De- cision-Making. Journal of Consumer Aﬀairs, 59(4), e70032. https://doi.org/10.1111/joca.70032 Khokhar, M. A., Anees, A., & Gujjar, M. W. (2025). From Judges to Algorithms: The Future of Corporate Adjudi- cation in AI-Driven Courts. The Critical Review of So- cial Sciences Studies, 3(2), 1794-1811. https://thecrsss. com/index.php/Journal/about Kumar, A. B., Sanjaya, K., & Saleem, M. (2025). Justice in the age of algorithms: Ensuring transparency, account- ability, and fairness in AI-driven legal systems. In C. Kaunert, A. Raghav, K. Ravesangar, B. Singh, & B. Riswandi (Eds.), Artiﬁcial intelligence in peace, justice, and strong institutions (pp. 191-210). IGI Global Sci- entiﬁc Publishing. https://doi.org/10.4018/979-8-3693- 9395-6.ch009 Mugamba, E. (2025). Digital Due Process in the Age of Al- gorithms: Redesigning Fairness in Criminal Justice Decision Systems in Uganda. Available SSRN. https:// dx.doi.org/10.2139/ssrn.6052435 Neil, R., & Zanger-Tishler, M. (2025). Algorithmic Bias in Criminal Risk Assessment: The Consequences of Racial Diﬀerences in Arrest as a Measure of Crime. Annual Review Criminology, 8, 97-119. https://doi. org/10.1146/annurev-criminol-022422-125019 Novelli, C., Floridi, L., Larsson, S., Taddeo, M., & Winter, S. L. (2026). The artiﬁcial in artiﬁcial intelligence how: imagination shapes AI regulation. Available SSRN. https://doi.org/10.1007/s43681-024-00434-4 Romeo, G., & Conti, D. (2026). Exploring automation bias in human AI collaboration a review and implications for explainable AI. AI & Society, 41, 259–278. https://doi. org/10.1007/s00146-025-02422-7 Rosbach, E., Ammeling, J., Ganz, J., Bertram, C. A., Con- rad, T., Riener, A., & Aubreville, M. (2026). Stuck on suggestions Automation bias the anchoring eﬀect and the factors that shape them in computational pathology. Machine Learning for Biomedical Imaging, 2026(Spe- cial Issue), 126-147. https://doi.org/10.48550/arX- iv.2603.11821 Selbst, A. D. (2025). Artiﬁcial intelligence and the discrim-

J. Law Epistemic Stud. (January - June) 4: e167 ination injury. UCLA School of Law, Public Law Re- search Paper No. 25-10; forthcoming in Florida Law Review. Available SSRN. https://doi.org/10.2139/ ssrn.5179224 Sivasatyanarayanareddy, M. (2024). Empowering access to justice through automated legal case management. International Journal for Multidisciplinary Research, 6(6), 1–13. https://philarchive.org/rec/SIVEAT UNESCO. (2021). Recommendation on the Ethics of Ar- tiﬁcial Intelligence. https://unesdoc.unesco.org/ ark:/48223/pf0000381137 Wenzelburger, G., Yeung, K., & Hartmann, K. (2025). Smart justice: Making sense of the rise of algorithm-based pre-trial risk assessment in criminal justice through legal models. Digital Society, 4(48). https://doi.org/10.1007/ s44206-025-00194-7 Conﬂicts of interest The authors declare that they have no conﬂicts of interest. Author contributions Ariadna Moya- Rodríguez and Yonayka Licea-Suárez: Conceptualization, data curation, formal analysis, investi- gation, methodology, supervision, validation, visualization, drafting the original manuscript and writing, review, and editing. Data availability statement The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Statement on the use of AI The authors acknowledge the use of generative AI and AI-assisted technologies to improve the readability and cla- rity of the article. Disclaimer/Editor’s note The statements, opinions, and data contained in all publi- cations are solely those of the individual authors and con- tributors and not of Journal of Law and Epistemic Studies. Journal of Law and Epistemic Studies and/or the editors disclaim any responsibility for any injury to people or pro- perty resulting from any ideas, methods, instructions, or pro- ducts mentioned in the content.