SUMMARY

This essay addresses the strategic and essential relevance of the constitutional right to understanding the decisions of public authorities, resulting from the consolidation and maturation of the constitutional principles of publicity, transparency, justification for administrative and jurisdictional acts, substantial due process, equality, adversarial proceedings, full defense, human dignity, proportionality, reasonableness, and other fundamental rights enshrined in the 1988 Constitution. The application of these rights depends vitally, in the digital age of complexity, on systemic traceability structured in intelligent, digital, auditable legal databases based on statistics and statistical models. In this sense, this work addresses the concept of the right to understanding as a systemic and deeper evolution, derived from the set of fundamental rights, which modifies the traditional paradigm of publicity and transparency in public sector decision-making bodies—especially in light of the mandatory observance of judicial precedents by judges, courts, and administrative authorities. In this regard, the aim is to contextualize the right to understanding within the normative universe that enables people's access to structured legal databases, artificial intelligence, statistics, and the auditability of algorithms and databases. At the same time, this essay seeks to demonstrate the relevance of these contemporary tools for understanding the decision-making acts of public authorities.

Finally, it aims to emphasize the importance of strengthening the culture, coherence, traceability, predictability, and self-criticism of authorities regarding the content and identification of decision-making patterns, as well as strengthening the culture of teaching, the culture of precedents, artificial intelligence, and access to databases in institutions as a whole — including education institutions and access to justice.

KEYWORDS

Right to Understanding; Precedents; Transparency; Publicity; Artificial Intelligence; Legal Databases; Statistics; Auditability; Public Decisions and Due Process of Law.

ABSTRACT

This essay addresses the strategic and essential relevance of the constitutional right to understand public authorities' decisions, as a result of the consolidation and maturation of constitutional principles such as publicity, transparency, reasoning of administrative and judicial acts, substantive due process of law, equality, adversarial proceedings, broad defense, human dignity, proportionality, reasonableness, and other fundamental rights enshrined in the 1988 Brazilian Constitution. In the digital era of complexity, the effective application of these rights depends fundamentally on systemic traceability, structured through intelligent, digital, auditable legal databases supported by statistics and statistical models. In this sense, the work discusses the concept of the right to understanding as a systemic and more profound evolution derived from the set of fundamental rights, modifying the traditional paradigm of publicity and transparency in public decision-making bodies—especially in light of the mandatory observance of judicial precedents by judges, courts, and administrative authorities. Accordingly, this essay aims to contextualize the right to understanding within the normative framework that enables public access to structured legal databases, artificial intelligence, statistics, and the auditability of algorithms and databases. Simultaneously, it seeks to demonstrate the importance of these contemporary tools in understanding public decision making acts. Finally, this essay also emphasizes the importance of strengthening a culture of coherence, traceability, predictability, and institutional self-criticism regarding the content and identification of decision-making patterns, as well as reinforcing a culture of education, precedents, artificial intelligence, and data access across institutions—from educational frameworks to mechanisms of access to justice.

KEYWORDS

Right to Understand; Precedents; Transparency; Advertising; Artificial Intelligence; Legal Databases; Statistics; auditability; Public Decisions; Due Process of Law.

1. Introduction

The classic right to publicity and transparency of judicial and administrative acts, provided for in Articles 5, items LX and XXXIII, 93, IX, and 37, caput, all of the 1988 Constitution, derives from the framework of contemporary liberal democracies. These same requirements of transparency and publicity coexist with the possibility of protecting intimacy, privacy, and confidentiality, whether in the cases provided for in the Constitution or in cases provided for by law 2. However, the digital age, intertwined with the concepts inherent to the era of complexity, in which transformations and the speed of events and plural thoughts interconnect, demands a rethinking of the hermeneutics regarding the scope of transparency and publicity surrounding decisions restricting fundamental rights. In the Brazilian Constitution of 1988, there is no doubt that it is necessary to interpret in a coherent and harmonious manner the requirements of transparency (art. 5, LX and XXXIII; art. 37, caput), publicity (art. 93, IX) and justification for judicial and administrative decisions (art. 93, IX; art. 37, caput), together with the mandatory observance of the prohibition of arbitrariness of public powers, resulting from substantive due process of law (art. 5, LIV), obedience to substantive and formal due process of law (art. 5, LIV and LV), compliance with legal certainty (art. 5, caput and XXXVI), equality (art. 5, caput and I), adversarial proceedings (art. 5, LV), full defense (art. 5, LV) and respect for human dignity (art. 1, III).

As if this indispensable integration of this set of constitutional requirements connected to the transparency and publicity of state decisions were not enough, the duty of coherence, objective good faith and institutional loyalty was also embraced by the legislator, under the aegis of the democratic principle, when providing for the mandatory observance of precedents by the Judiciary and administrative authorities, when issuing their decisions and formatting jurisprudence in the application of laws (articles 926, 927, 928, 489, § 1, items V and VI, and 1,036 to 1,041 of the 2015 Code of Civil Procedure).

In this context, the problem of judicial congestion, slowness, and overload affecting the judiciary is nothing new in Brazil. Furthermore, there is another serious structural problem: the unpredictability that plagues the system, given the lack of a culture of judicial precedent formation. The system adopted by articles 926, 927, 928, 489, §1, items V and VI, and 1036 to 1041 of the 2015 Code of Civil Procedure was not accompanied by a corresponding cultural implementation effort in the areas of education, training of judges, lawyers, and members of institutions essential to justice, much less by a national mobilization to foster this new culture.

There is no doubt that there is high-quality national literature on the theory of precedents, originating in English law and embodied, with adaptations, in North American law, from where it emerged into Brazilian law 3. Nevertheless, the Romano-Germanic culture of which Brazil is heir does not automatically adapt to a distinct and secular culture, which presupposes an entire tradition in the construction of precedents, and obstacles persist. In this essay, our purpose is not to address the theoretical obstacles, typical of the field of the great scholars of civil procedural law, related to the dogmatics of precedent theory.

As we warned initially, one of the first obstacles we will address in this work is related to the culture of teaching. However, there are essential pragmatic difficulties connected to the implementation of structured databases, artificial intelligence tools, statistics, transparency, and integration with the accessibility of administrative jurisprudence of institutions essential to justice, in addition to institutions that provide public services and enforce administrative norms. In short, the delivery of justice is not an exclusive prerogative of the Judiciary, and the constitutional principles that govern public administration, enshrined in Article 37, caput, of The 1988 Constitution must gain substantial depth to allow access to justice as a genuine fundamental right of the people, a circumstance that presupposes access to structured databases, closely interconnected with advanced artificial intelligence and statistical technologies, so that an authentic theory of judicial precedents can be constructed and administrative jurisprudence linked to these precedents, in a manner that is accessible and transparent to the public.

Furthermore, this essay also proposes to demonstrate that this new culture will strengthen institutions and the market with new paradigms of institutional integrity, quality, and protection of fundamental rights, as well as efficiency and competitiveness. In this scenario, our conclusion is that this new framework will be made possible through new transformative compliance models, both in the public and private sectors.

Regarding technological innovation in the legal field, it is important to reflect on three priority areas of action: (a) the careful implementation of artificial intelligence in the organization and analysis of documentary and case law collections; (b) systematic education on precedents in higher education institutions and government entities; and (c) expanding the application of the logic of precedents beyond the Judiciary, also encompassing the administrative sphere, such as audit courts, public defenders' offices, and regulatory agencies.

One of the hallmarks of the contemporary world, especially in Brazil, is the normative tangle and the constant profusion of laws, constitutional amendments, sub-legal normative acts, regulations, and rules of all kinds—a phenomenon that constitutes a permanently complex and continually changing normative network. As if this sophisticated machinery were not enough, this entire abstract normative apparatus undergoes a surprising metamorphosis when applied to concrete cases submitted for judgment in judicial and administrative instances, where the most diverse authorities boast decision-making autonomy.

In practice, we are talking about thousands of judges, appellate judges, and many other ministers of higher courts, as well as members of public prosecutors, public attorneys' offices, regulatory agencies, audit courts, autonomous agencies, decentralized administrations, and multiple state agencies, or even institutions overseeing activities essential to justice. The abstract normative tangle transforms into an even more complex and unpredictable jurisprudential multiplicity, aggravated by the difficulty of access for those administered and under their jurisdiction. This scenario greatly accentuates the compromise of expectations related to legal certainty, equality, transparency, impartiality, substantial publicity and prohibition of arbitrariness by public authorities.

2. Decisional transparency and databases as a constitutional imperative inherent to the right to understand decisions restricting fundamental rights

2.1. Legal Database: Architecture of Transparency and State Accountability

The State's actions, in their contemporary form, transcend the limits of formal legislation and judicial decisions. The legal binding of a person—whether human or legal—does not derive solely from rules developed by the Legislative Branch or decisions handed down by courts. They unfold through a multitude of legally binding State actions: administrative decisions, judicial and administrative jurisprudence, agreements concluded with public authorities, administrative contracts, sub-legal normative acts, and other administrative acts that, directly or indirectly, restrict, modulate, or recognize fundamental rights.

This plurality of decision-making, dispersed across different spheres and instances of public power, demands an institutional response commensurate with its complexity and impact. It is not enough for such manifestations to be formally accessible; it is essential that they be organized in a systematic, intelligible, and technically structured manner. Thus, the need for a legal database arises, not as a merely archival instrument, but as a material foundation for institutional transparency and state accountability.

This database should gather and make accessible the State's decision-making acts that, even if they do not have a typical normative form, produce relevant legal effects on the sphere of human rights. It is an infrastructure focused on qualified publicity, structured research, institutional auditing, and democratic governance.

In this scenario, it is imperative that we conceptually address the legal database and its impact on redefining the understanding of public authorities' decisions.

A legal database is, initially, a digital structure, whether public or private, but necessarily organized, intelligent and auditable, whose purpose is to capture, gather, receive, classify, order, explain, enable third-party interaction and make data and information intelligible, in order to optimize the institutional performance of the respective holder of this data.

bank and its users, respecting the fundamental and individual rights involved, preserving, when necessary, the limits inherent to the duties of confidentiality, in addition to the institutional memory of the decisions and patterns detected.

This definition of a database, which always involves legal aspects, recognizes the database as a necessarily intelligent entity. In this context, the legal database must perform at least some essential, fully auditable functions: ordering and classification; intelligence and statistical functions; interactive and organizational functions; and institutional and protective security functions 4.

This scope includes judicial decisions, administrative decisions,

judicial and administrative jurisprudence, judicial and extrajudicial agreements signed before or with public authorities, administrative contracts, normative acts and any administrative acts that produce legal effects on the sphere of freedom, property, self- determination or legal prerogatives of the person.

The legal database, in this sense, is not just an informational repository, but a technical-normative instrument aimed at consolidating public integrity, institutional predictability and social control over state acts.

We will examine the legal, constitutional, and technical foundations of the proposed concept, as well as the operational challenges and potential of its application in the context of the digital transformation of the State and the consolidation of data-driven governance models.

The function of a database in the legal field transcends the instrumental concept of a mere document repository. It is an institutional infrastructure oriented toward the systematization, rationalization, and transparency of legal knowledge, which carefully gathers and organizes judicial decisions, administrative decisions, case law, formal agreements, and administrative acts with significant legal impact. By adopting logical, chronological, thematic, and functional criteria, this type of database provides strategic support to interpretation and application of the law, allowing qualified access to precedents, normative foundations and coherent lines of argument.

When well-structured, a legal database directly contributes to promoting legal certainty, institutional predictability, and the effectiveness of justice. Its function is not limited to consultation: it acts as a tool for consolidating understandings, supporting legal research, and reinforcing the integrity of state decisions. Thus, it ceases to be a secondary technical instrument and establishes itself as a fundamental pillar in the architecture of institutional trust, especially in a context of increasing regulatory complexity and the need for public oversight of state actions.

2.2. Artificial Intelligence and Statistics as Infrastructure for Institutional Inference and Public Motivation

Statistics, in the era of mass, standardized decisions and institutional and technological complexity 5, should be understood as the scientific form of rational listening. It is a structured system of inference about regularities and exceptions, capable of identifying patterns, indicating risks of arbitrariness, and offering epistemic support for the legitimacy of public decisions. By transforming data into judgments, statistics allows the State to understand itself, revise its language, and act with predictability, responsibility, and prudence.

It is not merely a technical tool for quantification, but a formal language of reasonableness. It organizes the relationship between variability and coherence, offering objective criteria for distinguishing acceptable fluctuations from unjustified deviations. In a scenario where public decisions produce massive, immediate, and cross-cutting effects, statistics become a basis for structural accountability: it provides judges, managers, and regulators with a methodical mirror of the institution itself.

It is in this same context that artificial intelligence should be understood, especially in its predictive and explainable aspects. Artificial intelligence is not a substitute for public reason, but a technical extension of its analytical capacity. When guided by structured legal data and combined with statistical inference, AI can detect patterns of institutional behavior, recognize decisions outside expected parameters, suggest relevant precedents, and reinforce argumentative consistency 6.

Artificial intelligence acts as an instrument of interpretative traceability, allowing public motivation to be transformed from a formal or rhetorical gesture into a reconstructible, auditable, and comparable process. Its function is to broaden the scope of institutional attention, detect inconsistencies before they consolidate as structural ambiguity, and provide technical support for normative coherence. Like statistics, artificial intelligence doesn't decide: it illuminates, signals, and suggests—so that human judgment can act with greater depth, context, and prudence.

Integrated, statistics and artificial intelligence become the infrastructure for public motivation, institutional traceability, and the prohibition of arbitrary action. They operate as invisible pillars of a new form of decision-making accountability, which is no longer supported solely by the authority of the function, but by the verifiable coherence of its foundations.

Ultimately, it is about equipping public discourse with technical tools that reinforce its commitment to legality, predictability, and integrity in the 21st century.

Statistics and artificial intelligence in the context of public institutions in the 21st century XXI, must be understood as convergent expressions of the same applied rationality: institutional inference under uncertainty. Both operate not only on data; they operate on doubts, asymmetries, variations, and repetitions—those things that, in the daily grind of public decision-making, require prudence, comparison, and motivation. Statistics provides the method of listening. Rational; artificial intelligence expands the scale, speed, and capacity of pattern recognition. Together, they structure a silent verification architecture.

Statistics is not just a measurement technique. It is a scientific way of interpreting regularities, recognizing exceptions, and estimating risks based on evidence. It transforms dispersion into structure, variation into signal, and noise into diagnosis. In a complex institutional environment, marked by repeated decisions and conflicting interpretations, statistics act as a filter for reasonableness: it allows us to distinguish between legitimate variations and unjustified deviations. Its function is to anchor public discourse in criteria that can be audited, compared, and eventually revised.

Artificial intelligence, in its predictive and explainable aspects, should be understood as a computational continuation of statistical inference. Every model that suggests, classifies, or anticipates the outcome of an institutional decision does so based on probabilistic structures.

—sometimes hidden, but always inferential. When combined with structured legal databases and guided by traceability principles, AI becomes an interpretative extension of institutional memory: it allows us to identify relevant precedents, suggest argumentative convergences, and flag decisions that deviate from recognizable norms.

Public motivation — which, at the legal level, requires clear and verifiable grounds—finds legitimate technical support in statistics and artificial intelligence. It's not about replacing judgment, but about qualifying it. Decisions that incorporate statistical inference and computational intelligence are no less human; they are more nuanced, more contextualized, and more open to public criticism. Motivation ceases to be a ritual of language and becomes a manifestation of institutional coherence fueled by standards, references, and accountability.

By integrating statistics and artificial intelligence, the State reinforces its commitment to the traceability of decision-making and the prohibition of arbitrariness. Where there are identifiable patterns, there must be criteria to justify ruptures. Where there is normative regularity, there must be control over exceptions. The role of these technologies is not to decide—it is to illuminate. They are tools for institutional listening: they allow the State to listen to itself, compare itself, explain itself, and, when necessary, correct itself.

This understanding — that statistics and artificial intelligence, integrated and based on structured legal databases, should act as the technical infrastructure for public motivation and as guarantees of traceability, coherence and the prohibition of arbitrariness — was recently corroborated, on an international scale, by the study of

Chutisant Kerdvibulvech (Big Data and AI-driven evidence analysis: a global perspective on citation trends, accessibility, and future research in legal applications, 2024) 7.

Kerdvibulvech demonstrates, through empirical analysis and a global literature review, that artificial intelligence systems applied to legal analysis — especially in document review, litigation prediction, forensic image analysis, and contract evaluation — only produce legitimate and admissible effects when accompanied by rigorous statistical validation, methodological traceability, and transparent ethical parameters. Statistics, in this context, do not appear as an accessory technique, but as an epistemic guarantee of institutional rationality.

Kerdvibulvech argues that statistical inference is essential for controlling biases, measuring uncertainty, and identifying patterns and exceptions. At the same time, artificial intelligence must be applied under interpretable and auditable guidelines, so that institutional decisions—administrative, judicial, or investigative—do not become automatic gestures, but rather motivated acts with depth, prudence, and inferential responsibility.

This finding confirms the central thesis of this topic: contemporary public decision- making requires, in addition to legal grounds, a technical basis for inference, verification, and explanation to enable the right to understanding. Furthermore, decisions need to be interpreted within a systemic context to enable societal understanding. By integrating statistics and AI, the public institution commits to a systemic and complex decision-making language model, authentically integrated, capable of resisting structural ambiguity, preventing normative inconsistencies, and ensuring predictability without rigidity. Motivation ceases to be a formal requirement and becomes a continuous exercise of institutional listening: listening to data, patterns, ruptures, and the limits of one's own decision-making power.

The structuring of case law, administrative, and business databases should not only serve retrospective statistics or predictive artificial intelligence. Their deeper role is to provide a stable, auditable, and technically grounded language for the conclusion of out-of- court settlements and to guide public authorities in their decisions, whose legitimacy depends on consistency with past decisions—whether judicial, administrative, or business. The lack of this anchoring in precedents and prior agreements compromises not only fairness between parties in similar situations, but also the but also the logical integrity of the normative function exercised by institutions. As Chutisant Kerdvibulvech (2024) demonstrates, disorganized or disjointed data weaken artificial intelligence systems, impede pattern detection, and obscure systemic deviations. Therefore, to be legitimate, effective, and transparent, judicial or extrajudicial agreements must be integrated into the institutional memory formalized within contemporary technological standards rather than 20th-century methodology, reflecting already recognized standards and allowing public control over their consistency with the historical language of decisions.

In this context, the coordinated application of statistics and artificial intelligence offers public and private institutions the possibility of developing a cognitive infrastructure focused on analyzing, cross-referencing, and validating these databases. Statistics allow for mapping decision-making frequencies, identifying recurring argumentative patterns, and recognizing hermeneutical inflection points—including in historical series of out-of-court settlements and administrative decisions. Artificial intelligence, fueled by this structured universe, becomes capable of performing more sophisticated tasks: detecting internal inconsistencies, pointing out unjustified divergences between similar cases, assessing adherence to precedents, and predictively flagging risks arising from solutions outside the institutional framework.

This combined functionality acts as a silent engine of coherence. Systems trained with properly classified, versioned, and traceable data can offer, for example, suggestions for clauses aligned with the terms of previous similar agreements, alert to the risks of contradictory decisions, or project regulatory impacts not yet perceived by traditional legal rationality. The potential of this analytical architecture is not limited to efficiency, but reaches a deeper level: institutional self-awareness. It allows organizations to observe themselves, review themselves, learn from their own records, and, above all, establish a language that can be recognized and replicated responsibly.

The constitutional right to understanding inaugurates a new interpretative paradigm and breaks the limits of administrative and jurisdictional transparency and publicity. Decision-making acts must be understandable; it is not enough to be public and transparent. They must be well- founded, rational, and coherent. No one understands an arbitrary act. Coherence ceases to be merely a rhetorical aspiration and becomes monitorable, auditable, and measurable by metrics informed by empirical standards. Even an out-of-court or judicial settlement ceases to be an isolated gesture of convenience and becomes part of an ecosystem of decisions.

Interdependent, subject to technical scrutiny and comparative review. This expands the capacity of the State and corporations to engage with their own precedents—judicial, administrative, and business—without losing sight of the uniqueness of each case. Artificial intelligence and statistics, together, do not replace institutional deliberation, but rather accompany it with an ethical horizon: avoiding distortions, preserving memory, and preventing arbitrariness from masquerading as discretion.

For the theory of precedents to operate as the rational core of the legal system—as required by stare decisis—it is essential that case law be analyzed not only hermeneutically but also through statistical and computational resources capable of diagnosing its internal fractures. Statistics, in this field, allows us to identify divergent patterns between decisions on analogous cases, map the dispersion of reasoning across different chambers and courts, and quantify the degree to which decisions adhere to or deviate from qualified precedents, signaling which areas offer legitimate controversy or hermeneutical discretion and which areas are pure anomalies and arbitrariness. This is a diagnostic and predictive function: it highlights where the system behaves coherently and where, due to interpretative or contextual flaws, it begins to lose its consistency and even enter suspect zones.

Artificial intelligence, powered by structured databases with robust metadata (topic, thesis, adjudicating body, rapporteur, outcome, main grounds, legal provisions invoked), can go further: it can not only identify these inconsistencies but also project future deviations based on emerging decision-making trends. Through supervised algorithms, it is possible to train models that indicate, for example, the likelihood of a given thesis being revised, challenged, or ignored by certain instances or regions. This monitoring is vital for preserving the integrity of the precedent system, as it allows not only early warning of the erosion of consolidated understandings but also the continuous calibration of judicial language based on its own decisional memory.

For this process to be reliable, however, the databases' feed is crucial. Without adequate technical curation, artificial intelligence becomes blind and statistics become illusory. It is essential that decisions are correctly classified, that qualified precedents are clearly marked, and that there is an institutional protocol for recording relevant grounds. The data needs to be cleaned, updated, standardized, and enriched with context. It's not just about digitizing judgments: it's about necessary to transform decisions into structured language, with specific fields that allow their algorithmic interpretation without loss of legal density. This structure will allow us to filter relevant cases, distinguishratio decidendiofobiter dicta, and accurately map the cores of normative meaning that radiate from the precedents.

Thus, the preservation of stare decisis in the age of complexity will not be guaranteed solely by formal declarations of binding force, but by the institutional capacity to systematically monitor, audit, and project legal coherence. Statistics and AI, in this sense, operate as instruments of lucidity: they reveal the structure behind discourse, the regularity behind exceptions, and the instability behind the appearance of uniformity. Law ceases to be a self-centered narrative and becomes a field of empirical observation and interpretative responsibility.

The logic of structuring, statistical analysis, and predictive interpretation applied to case law and out-of-court settlements should be extended to public contracts and regulatory acts, whose regulatory effects are often broader and more lasting than specific court decisions. The creation of structured databases containing contractual clauses, performance conditions, addenda, legal opinions, and practical results allows artificial intelligence to detect abusive recurrences, strategic omissions, inconsistencies in interpretation, and asymmetries in treatment between different contracting parties in similar situations. In the regulatory field, the systematic organization of resolutions, instructions, ordinances, and decrees, with metadata on legal basis, issuing agencies, express motivation, and validity, enables analyses that reveal the degree of interpretative uniformity between federative entities and regulatory agencies. The statistics applied to this set become a tool for continuous constitutional auditing, capable of verifying whether normative acts comply with legal frameworks and align with the system of precedents, avoiding contradictions, normative redundancies, or regulatory gaps. When fed with technical rigor and processed by explainable AI, these collections become part of an integrated institutional intelligence, in which contracts and norms not only produce legal effects but also feed back into the state's normative memory, allowing public governance to learn from its own actions and anticipate recurring deviations

2.3. Artificial intelligence, traceability and prohibition of arbitrariness

The application of artificial intelligence in public administration is a topic of growing importance. In this context, the relevance of this technology in systematizing and categorizing the reasons underlying the decisions of different agencies in similar situations stands out. This organizational capacity not only allows for the identification of inconsistencies but also helps clarify interpretative inconsistencies. Furthermore, systematization favors the promotion of institutional consistency, which is highly desirable.

However, it is essential that the implementation of artificial intelligence in this area is guided by public and auditable criteria 8, respecting the constitutional principles in force. This guidance is especially important regarding the principles that guarantee legal certainty and adequate justification for administrative acts. Motivation, understood as the requirement of a rational basis, is closely related to the concept of traceability. An administrative act that presents adequate motivation is one whose reasons can be easily reconstructed, verified, and challenged.

It is worth noting that this requirement of motivation is not limited to judicial decisions. It also extends to administrative sanctioning acts, binding opinions, regulatory resolutions, and agreements signed between public entities. In this sense, the organization of administrative jurisprudence in structured, improved databases.

Powered by artificial intelligence and made publicly available, this represents an effective strategy. This approach not only ensures transparency but also establishes a model of administrative justice based on the coherence, rationality, and integrity of legal language.

Finally, it is imperative that technological advances in the public sector be accompanied by a firm commitment to the values that underpin the democratic rule of law. The responsible integration of artificial intelligence can thus significantly contribute to a more efficient, fair, and transparent public administration.

2.4. Integrative hermeneutics of constitutional law to understanding:

convergence between administrative and judicial jurisprudence through databases

The constitutional right to understand the content of public decisions restricting fundamental rights requires a hermeneutics that views the set of decisions as a whole—that is, a normative system properly structured, organized, classified, and capable of in-depth research in the technological age. In this sense, this hermeneutics, regardless of the current it purports to designate, is based on an unavoidable contemporary assumption: viewing the decision in its comprehensive and organized normative context. It is impossible to ignore that this context is part of the age of complexity, as has been stated from the outset 9.

Statistics, in its contemporary approach, transcends the mere measurement of quantitative phenomena. It establishes itself as a science that encompasses structure, inference, and decision-making, focusing on organizing information, identifying patterns, anticipating risks, and providing a rational basis for institutional choices.

legal context, this function acquires crucial importance, proving to be a vital element for public rationality, consistency in decisions and the integrity of governance.

Legal statistics are based on three fundamental pillars: first, the systematic and structured collection of relevant public data; second, the mathematical modeling of identifiable patterns in different contexts, ranging from regulatory to judicial; and finally, responsible inference, which must be auditable and publicly justifiable, regarding risks, trends, repetitions, and deviations. Thus, its object of study goes beyond numbers, encompassing institutional behavior, the language of decision-making, and the logic of institutions in uncertain situations.

In this sense, statistics transform the legal database into a space for systemic observation, enabling the identification of asymmetries, the prediction of conflicts, and the rationalization of state action. The relationship between statistics, databases, and legal language is, therefore, structural in nature. Without reliable, organized, and auditable data, legitimate inferences cannot be made; similarly, without institutionalized statistics, databases become mere technical repositories, devoid of analytical value.

Furthermore, without a standardized legal language, classifying, cross-referencing, and interpreting decisions becomes a challenge.

Within the justice system, statistics perform four central functions. First, a diagnostic function, which seeks to identify interpretative patterns, areas of instability, inconsistencies in decisions, and unequal treatment in similar cases. The preventive function, in turn, anticipates the emergence of conflicts, legal risks, or argumentative distortions, based on historical data and institutional patterns. Next, the strategic function underpins the management of case provisioning, the prioritization of agendas, and the structuring of coherent public responses. Finally, the restorative function provides objective support for reviewing dysfunctional practices and correcting institutional biases, also encompassing the perspective of external oversight.

For these reasons, it is essential that statistics be understood as a principle of democratic governance, especially in complex and sensitive legal environments. It strengthens the state's capacity for introspection, allowing it to review its structures and promote actions based on standards of coherence, efficiency, and equity. The application of this rationality in the justice system is imperative in times of

complexity. To this end, courts, regulatory agencies, public prosecutors, audit courts, and internal control bodies must incorporate statistical tools guided by clear public purposes and ethical, auditable routines that are permanently aligned with the constitutional language.

Thus, legal statistics should not be seen merely as an auxiliary technique, but as an epistemological foundation that sustains institutional integrity. It does not replace argumentation, but enhances it; it does not supplant the norm, but rather the structure; and it does not automate justice, but rather anchors it in evidence, memory, and public accountability. By translating legal data into analytical language, statistics ground justice as an expression of collective intelligence.

3. Metric Transparency in the Age of Algorithms

3.1. The role of statistics in the legal world and in improving the justice system

It is important to note, from the outset, the conceptual difference between data and algorithms, although it is assumed that these concepts underlie the logic of this essay. The concepts taken fromArtificial Intelligence Risk Management Framework (AI RMF 1.0), prepared by the National Institute of Standards and Technology – NIST (2023), as well as the Regulation (EU) 2024/1689 of the European Parliament and of the Council, which deals with artificial intelligence within the European Union (EUROPEAN UNION, 2024). Both are based on essential premises inherent to digital governance 10.

The conceptual distinction between data, metadata, and algorithms constitutes a structuring element in contemporary debates on artificial intelligence, automated governance, and digital regulation. This distinction is essential for the proper interpretation of the obligations imposed on institutions seeking to ensure fair and equal access, comprehensible reading, and informational protection in the ethical use of data within the Brazilian legal, economic, and communication space—especially in digital environments such as social media, where data processing reaches massive proportions and amplified effects.

Data are digital records or representations—structured or unstructured—of any fact, act, occurrence, state, process, or event that can be captured and has relevance. Metadata, in turn, is data about the data, according to the context of its collection. In the European legal and technical context, it encompasses both raw data (such as the date and time of a purchase) and processed data (such as a user's consumption history). In the RFM model, data corresponds to the observable elements that feed the metrics: number of purchases made, date of the last transaction, and total amount spent.

Algorithms, on the other hand, are finite sets of logical or mathematical rules and instructions used to process data, extract patterns, categorize subjects, or make automated decisions. An algorithm based on the RFM model must be explainable through predefined formulas, based on risk models contextualized by segment.

From a regulatory perspective, European Regulation (EU) 2023/2854 classifies algorithms as operations on data that must comply with fundamental principles such as proportionality, non-discrimination, necessity, and transparency. This classification implies recognizing that algorithms are not immune to oversight, especially in contexts where they impact fundamental rights, access to essential services, or the classification of standards, agreements, and decisions that directly affect these rights.

In the Brazilian context, the General Data Protection Law (LGPD – Law No. 13,709/2018) establishes fundamental guidelines for the processing of personal data, highlighting the principles of purpose, adequacy, necessity, free access, data quality, transparency, security, prevention, and non-discrimination. The LGPD also imposes obligations regarding the protection of confidentiality and information security, requiring both the controller and the processor to adopt effective technical and administrative measures to safeguard personal data against unauthorized access or incidents—whether accidental or unlawful—that may result in the destruction, loss, modification, improper communication, or dissemination of such information.

The distinction between data, metadata, and algorithms, therefore, is not merely bureaucratic: it is a structural element for the ethical and normative governance of digital systems. Data constitutes the raw material; algorithms, the instruments of transformation. Protecting the integrity of the process requires oversight of both: the origin, classification, and use of data, as well as the criteria and impacts associated with algorithms.

The structuring of documentation, rigorous data processing, and the auditability of digital systems, including algorithms, provide new paradigms for the legitimacy of public decisions.

The consolidation and strengthening of precedents are connected to the right to understanding, which derives from a constitutional system integrated into a dynamic, self- critical, and regenerative institutional culture. This culture must not only defend itself but also constantly renew itself and engage in dialogue with the State. In this context, transformative compliance in companies also emerges, emerging as the link between the technical structure and the institutional essence, representing the ethical intelligence that guides artificial intelligence, the integrity that organizes data, and the trust that underpins predictability, with regulatory autonomy and a new perspective on the organization's identity.

As I mentioned, the complexity of decision-making, the lack of transparency, and the difficulty in accessing state decisions and acts stem from multiple factors. The legal database lacks integration with statistics, as it is an intelligent tool that enables a sophisticated reading of institutional memory. It has become clear that statistics reveal patterns and deviations, allow for the critical interpretation of trends, hidden flaws, fissures, inequalities, asymmetries, and diagnoses regarding normative and decision-making implications. In the current context, the improvement of statistical tools shows trends toward integration with artificial intelligence, thus fulfilling extremely important functions in the analysis, diagnosis, and anticipation of scenarios and risks in the institutional decision- making architecture. In these new scenarios, AI considerably expands the scale and scope of statistical impact, and technological advances increasingly enable the identification of invisible and sophisticated patterns. Methodologies exist that allow for continuous revisions as evidence emerges or undergoes transformation. In any case, it should be noted that statistics and artificial intelligence do not replace human judgment, but constitute auxiliary instruments and have auditable and variable methodologies 11.

3.2. How algorithms impact and structure legal databases

Regarding the concept of an algorithm in information technology, it's crucial to remember that it's a structured set of instructions designed to solve problems or perform tasks automatically. In the digital world, an algorithm operates like a recipe that guides computers in their decisions, based on data. Rather than adopting a random approach, the algorithm follows meticulously designed steps aimed at classifying, ordering, correlating, or predicting information. This dynamic proves crucial, for example, when an electronic legal system is able to identify analogous decisions, when a platform detects contractual risks, or even when a compliance program identifies inconsistencies in business operations.

Although developed by technology experts, algorithms are not neutral; on the contrary, they reflect human choices about what should be valued, what can be disregarded, and which paths deserve to be prioritized. Therefore, their application in legal spheres requires a commitment to accountability, transparency, and oversight. When properly designed and audited, algorithms have the ability to organize vast volumes of data, reduce the incidence of errors, and increase the consistency of decisions. However, when they operate without proper oversight or with biased data, they risk perpetuating inequalities, creating legal vulnerabilities, and compromising institutional integrity.

Algorithms currently constitute the dynamic axis that transforms legal databases into functional systems of institutional rationality. If the database is responsible for storing and organizing decisions, normative acts, contracts, and opinions, they are the algorithms that give form, intelligibility, and operability to this collection 12. Through them, it becomes possible to classify documents, extract linguistic and normative patterns, identify recurrences, assess argumentative coherence, and build inferences applicable to new cases or regulatory scenarios.

In contemporary legal databases, algorithms perform a variety of structural functions, such as semantic indexing and hierarchical organization of content, where classification algorithms categorically organize decisions according to subject matter, legal basis, adjudicating body, cited case law, type of request or legal thesis, ensuring accurate information retrieval, even in massive and heterogeneous collections.

Furthermore, detecting patterns and inconsistencies becomes a vital function, with clustering and anomaly detection algorithms facilitating the identification of inconsistencies between similar decisions, revealing areas of jurisprudential instability and locating interpretations outside the historical norm. Temporal traceability and predictive inference are also crucial, with time-series algorithms enabling the detection of decision-making seasonality, interpretative ruptures, and normative effects over time, which is essential for statistical inferences, resource allocation, and institutional impact simulations.

Explainability and legal inference in artificial intelligence are equally relevant; in more advanced systems, algorithms become logical inference mechanisms, used to suggest arguments, reconstruct legal grounds, predict procedural outcomes, or recommend contractual clauses, with all of this functionality dependent on the integrity of the database and the ethical governance of the applied algorithm.

Finally, filtering, cleaning, and anonymizing sensitive data through pre- processing algorithms is essential to eliminate duplication, correct inconsistencies, and ensure the anonymization of personal data before it is used for AI training or public analysis.

As a result, legal databases are transformed into algorithmic environments, ceasing to be mere passive repositories and becoming dynamic infrastructures for inference, institutional memory, and public or corporate intelligence. Without algorithms, there is only a mass of documents; with their application, there is structure, meaning, criticality, and decision-making potential. However, this transformation is only legitimate if the algorithms are: explainable, so that one can understand how they perform classifications, recommendations, or exclusions; auditable, so that