Judicial independence and the use of artificial intelligence in courts

DOI 10.51634/2307-5201_2023_2_15

УДК 347.91

ГРНТИ 10.31.01; 10.33.01

D. Drozd,

Attorney, independent researcher

(Izmir, Turkey),

e-mail: Адрес электронной почты защищен от спам-ботов. Для просмотра адреса в вашем браузере должен быть включен Javascript.

The relevance of the topic of this research lies in the lack of legal papers on the questions of using computer technologies in judicial proceedings. As artificial intelligence (AI) systems continue to advance, there is a growing need to address their potential impact on judicial principles. Various countries try to implement AI technologies in judicial dispute resolution, these processes are progressing intensively, but there is still a significant gap in the existing legal literature on this topic.

The subject of this paper is the impact of AI systems’ use in courts on judicial independence. The purpose of the study is to examine the implications of AI technologies on judicial independence and explore how adherence to specific criteria can help maintain the autonomy of the judiciary in the digital age.

The topic of this research is highly novel due to the nonexistence of legal papers dedicated to the problem of this study. While many scholars are focused on judicial independence, considering it a crucial principle for the rule of law, there is a growing number of researchers analyzing the implementation of AI technologies in court. However, these scientific works primarily concentrate on the ways of possible using AI systems in courts and other general questions. However, there are only a few papers devoted to the analysis of legal risks associated with the use of AI systems in courts. This work is the first one that opens the door to further discussions on the assessment of judicial independence in AI-assisted proceedings.

The research methods used in this paper are typical for legal studies. By analyzing the relationship between AI technologies and judicial independence, this research seeks to shed light on the potential challenges that arise from the integration of such computer systems into judicial systems. The current work addresses different approaches to defining key features of judicial independence and then proposes a list of specific criteria necessary to ensure judicial independence when utilizing AI systems.

The key conclusions of this research contribute to the scholarly discussion on judicial independence criteria and provide two new specific criteria for AI-assisted proceedings. The author suggests that adhering to this list of specific criteria can provide independence for AI systems themselves and for judicial systems in general. By exploring the criteria for safeguarding judicial independence in the context of implementing AI technologies, this article aims to offer valuable insights and recommendations for policymakers and stakeholders.

Keywords: artificial intelligence; judicial independence; AI in courts; digital courts; independence of courts.

Introduction

Almost every sphere of our day-to-day life is subject to constant changes due to digitalization. The judicial sphere is not an exception. Enthusiasts from all over the globe launch various projects in order to implement artificial intelligence technologies into judicial decision-making. Research by Oliveira, Silva Gomes, Enes et al. suggests, while studies on the public use of artificial intelligence (AI) generally are on the rise, there is still a lack of such research papers in the field of justice.

Digital technologies help judicial systems operate more efficiently and be more convenient for the parties to a case. It is undeniable that videoconferences, for instance, make participation in court proceedings cheaper and easier especially when the parties of a dispute are located remotely from a court. However, not every modern technology brings only positive effects to judicial proceedings. While AI technologies have been integrated into the judicial sphere in various countries during the last decade, there are many concerns on this issue, which are not sufficiently covered in legal papers. This article aims to analyze the fate of judicial independence in the digital age and determine whether it is possible to secure judicial independence in the case of using AI systems to decide cases in courts or to assist judges to do so.

AI systems may be used in legal proceedings in various ways. Firstly, courts can use these technologies to assist judges in finding proper precedents, checking whether legal acts and precedents were abolished or amended, checking the correctness of parties’ legal positions from the point of view of actual court practice, translating written evidence and oral statements from foreign languages, evaluating evidence, preparing the drafts of judicial acts and so on. Secondly, when it comes to deciding cases by AI systems, these systems may be empowered to partly replace judges in easy cases in the courts of first instance. Such “computer-judges” are intended to analyze the arguments and evidence of the parties to a case, then reach a decision, and prepare its text in a human-readable format. So, there are numerous possible ways to use AI in courts, but in this article, we will explore the question of judicial independence in general without separation on different AI system types.

This paper begins by establishing the need to define the concept of judicial independence. After that, we analyze the list of criteria that help to assess the level of judicial independence in a specific jurisdiction. Then we rewrite these criteria to make them applicable to AI-assisted proceedings. Finally, we propose two additional criteria which are needed to safeguard judicial independence with the use of AI systems.

Main Provisions

Starting with the role of judicial independence, we revealed that the most important effect of this principle is the protection of judicial systems from undue influence by the government. Other branches of a government should not influence the work of judges, so the principle guarantees that judges cannot be coerced by powerful political actors to make decisions against their will and their understanding of proper judicial proceedings. Meanwhile, judicial independence also protects courts from other actors who may have interests in a particular case; however, the majority of scholars emphasize political protection as a primary objective. Thus, the definition of judicial independence demonstrates that this judicial principle safeguards judges from any external influence and provides them with confidence that they cannot be penalized for their decisions.

There are a number of criteria that enable to estimate whether a judicial system is independent or not. This paper provides an analysis of different approaches to this issue, and then proposes a list of the seven most valuable criteria of judicial independence. These criteria are simultaneously the steps that need to be taken in order to achieve judicial independence.

The criteria which were originally described for ordinary judicial proceedings are applicable to AI-assisted proceedings as well, but with slight adaptation. However, to safeguard judicial independence in courts equipped with AI systems, two more criteria should be considered in addition to the seven general criteria. These additional criteria designed by us aim to protect the independence of AI systems from judges, judicial clerks, and other specialists involved in creating and maintaining AI products for dispute resolution. Using nine mentioned criteria altogether shall constrain external influence on decision-making and enable judicial systems to operate properly and lawfully. As the author believes, such AI systems shall align with the definition of judicial independence, and their use will not jeopardize the principle of judicial independence.

Materials and Methodology

In this paper, the results of in-depth research by different scholars on the topic of judicial independence were used. These results were analysed, interpreted, and adapted in order to be applicable for AI-assisted proceedings. The author employed typical legal research methodology, which involved a systematic analysis of the concept of judicial independence. The initial step was to define and understand the principle of judicial independence and its role. The paper then focused on exploring the reasons why judges and courts should remain independent and who may seek to influence them.

Overall, the methodology of this paper involves a comprehensive examination of the various aspects of judicial independence, considering its role in the digital age, and its significance in protecting judges, courts, and AI systems.

Research Results

1. The Role of Judicial Independence

In order to analyze the destiny of judicial independence in the digital age, the primary task is to define this principle itself. The current part of the essay is devoted to the role of judicial independence.

When we think about the concept of the independence of judges, what do we usually envision as its foundation? Why should judges be independent? It seems obvious that the lack of dependency allows judges to make the right decisions (as long as judges sincerely believe in them) without fear about what someone would think or do after the announcement of a decision. In this context, it is essential to identify the actors – judicial independence is usually understood as independence from whom or what?

The main actors, who might be interested in influencing judges, can be categorized into the two main groups, namely a government and a society. A society includes the parties of a judicial dispute, and other private citizens and companies. In other words, these two groups can be named as “political pressures” and “other pressures”.

The independence from a corresponding government is of high importance and the most difficult to achieve. A fundamental work on judicial independence, written by Tom S. Clark, begins with a thesis that the separation of powers is the most significant feature of the Constitution of the USA and probably the most valuable contribution of it to the democracy throughout the world. Other scholars underline that the separation of powers enables judges to act fearlessly, independently, and effectively. The separation of courts from politics serves as a key to distinguish advanced countries and market-based democracies from failed democracies with a high level of corruption and from authoritarian regimes.

At the same time, even in the USA courts are partly dependent from other government bodies. Firstly, this dependency lays in the field of bringing judicial decisions into effect because executive powers choose whether and when to take actions to enforce judicial decisions. Secondly, the United States Congress controls at least the budget, the structure, and the jurisdiction of courts. However, the both cases are not the most dangerous ones. It is usually believed that such dependence between different branches of a government is a good thing that – if made wisely – creates a balanced system and prevents the tyrannical abuse of power.

The most harmful form of a government’s influence is direct orders to judges to decide a specific case in a specific way. This phenomenon is usually called as “telephone justice”. This way of disturbing judges is perfectly known in undemocratic societies, where judicial independence from politics is often more declarative rather than real. In such jurisdictions, judicial independence, even if it is mentioned in legal acts, is much more associated with independence only from a society and from the parties of disputes, in particular. 

Independence from ordinary people seems to be an easier task. This task includes safeguarding judges mainly from the following two factors: positive influence (corruption) and negative influence (criminal threats to judges, their relatives, and their property). We suppose that in countries with relatively low criminal rates, to the contrast to undeveloped countries, these threats can be almost eliminated by corresponding states, since the states have a monopoly on the legal use of force and the states are highly interested to be the ones who fully control the judicial system of a country.

At the meantime, independence both from political and other pressures has a critical importance due to a number of reasons. To begin with, judicial independence enables laymen to believe that their potential disputes can be resolved fairly. Public respect for the judicial system of a country motivates people, companies, and public actors to bring their disputes into the courts and resolve it in a civilized manner. That makes judicial independence a highly important feature of the rule of law. Next, the independence of the judicial branch of a government makes the regulations and the precedents of a corresponding jurisdiction more valuable. The decisions of independent judges are more predictable because they honestly rely on law familiar to the society. Otherwise, unfair courts and unpredictable regulations would lead market actors to protect themselves via other tools, making contract prices higher and decreasing the level of investment attractiveness. The last but not the least is that independent courts protect individual rights from governments which allegedly overstep their authority. Only independent judges are empowered to make a government be consistent with its nation’s laws, enabling regular citizens and opposition leaders to challenge the decisions and actions of government officials.

All in all, it seems that the main role of judicial independence is to provide a protection to judges from the other branches of a government. As we can see in various legal papers on this topic, the concepts of separation of powers and of judicial independence are quite often linked to each other, and the latter is usually described as the integral part of the former. Moreover, some scholars use the word “autonomy” as the contextual synonym of “independence” that leads us to the clear understanding of judicial independence as a political characteristic rather than a safety feature against regular citizens. And even the development of judicial independence is linked by some authors to political grounds like the will of ruling regimes to protect themselves after predicted turnover.

However, the function of this principle to safeguard judges from a society still exists to some degree because corruption and other threats from interested persons and powerful actors are not abolished.

2. The Definition of Judicial Independence

According to Michael D. Gilbert, judicial independence is a condition when judges cannot be penalized by other actors for their official decisions and judges know that they cannot be penalized. “Other actors” were identified by the author in his article widely, including political powers, interest groups, parties to a case, and even other judges, meaning also independence of individual judges, not only the independence of the judiciary in general. Speaking of judges’ knowledge on their safety, Irving R. Kaufman also underlined the importance of judges being assured that their decisions cannot “lead to personal punishment”, unless a serious offense was committed.

Professor Paul Gewirtz defines judicial independence as an ability to resolve disputes “in a lawful and impartial manner, free from” external influence and any improper control. A similar definition is given by Jeffrey L. Dunoff and Mark A. Pollack: the judges’ freedom to decide cases “free of outside influences”.

Bearing in mind that the majority of researches understand judicial independence in a political way, it seems reasonable to describe the phrase as “the ability and willingness of courts” to resolve disputes lawfully having no undue regard to the opinions of government actors. However, even the authors of this definition mentioned the existence of the other types of influence: lower courts are dependant, in some legal systems, from higher courts, particularly when the promotion of judges is controlled by senior judges; and courts may also be influenced by a corruption. At the same time, the independence of supreme courts has the greatest meaning since they are “the courts of last resort” in most of cases, and the independence of the entire judiciary is affected by the independence of higher courts. Thus, political interference is the most important one because political pressure is the main risk for the independency of higher courts (and still a risk for lower courts even if there are more dangerous pressure actors for them).

In legal literature, it is often distinguished formal (negative or, as it is called in other scientific papers, de jure) judicial independence and positive (de facto) independence, where the former is linked to the guarantees of independence, while the latter refers to judicial behavior. Formal independence includes the rules designed for the appointment of judges, their jurisdiction and discipline, tenure, and the budget of a judicial system. Negative independence does not automatically lead to gaining a positive one; positive independence can be assessed only in practice during adjudicating cases with political interests in it. 

To sum up, we share a common point of view that judicial independence is equally relevant for both the judiciary as a whole and individual judges as a part because both should be safe of “external pressures and of each other”. Thus, we propose to understand judicial independence as the state when courts and individual judges have the ability to decide cases free from any external pressure with a confidence in their safety from any type of punishment. This definition implies that both any actor around a specific judge has no power to somehow influence the decisions of that judge and any actor outside the specific court or judicial system in general has no power to influence decisions in specific cases or the course of resolving specific matters in general.

3. Judicial Independence Criteria

In this part, the list of human-judges’ independency guarantees and indicators will be mentioned in order to provide an opportunity then to decide which of them and in what manner might be relevant for machine-judges. 

Speaking about the possibility of political attacks, J. Melton and T. Ginsburg highlighted the following provisions: the formal statement of judicial independence; a lifetime tenure (or, at least, fixed tenure, as others declare); an appointment process with the involvement of several various actors or a judicial council; a difficult removal procedure (for example, a supermajority vote in a legislative body); the limited list of removal conditions; and salary insulation (or budgetary autonomy in general). We should notice that these guarantees protect the independence of judges not only from governments but also from ordinary people. For example, according to Alexander Hamilton, who criticized periodical appointments of judges, the commitment of appointing powers to the people of a corresponding state would not be better than committing it to either the executive branch of a government or legislature.

Next, the list of measurable criteria should be presented. For these purposes we will mix several approaches of different scholars with our own views. Accordingly, the criteria, which can be used for assessing the level of judicial independence, are the following:

• the independence of judges and the judicial structure are formally fixed in a legal act, which cannot be easily changed (e.g., in the constitution);
• the appointment of judges is not controlled by one powerful politician (e.g., the president, the prime minister, the minister of justice, or the president of the supreme court) and ideally involves professionals from different legal backgrounds or, at least, from the different branches of a government;
• a lifetime tenure or an appointment for up to a fixed retirement age without renewable terms (including for the presidency at the supreme court) and with a detailed provisions on situations of changing that age (e.g., whether it should be increased in the case of increasing ordinary civilians’ retirement age or not);
• a removal procedure is tough and relies on the limited list of grounds, individual powerful actors cannot remove judges (the same way as they cannot appoint them – see (2) above);
• the budget of the judiciary should be safe from possibly interested branches and the salaries of judges should not be reduced but should be corrected regularly due to the official rate of inflation (or interest-rate alternatively) and should stay adequate in comparison to other legal jobs;
• automatic or strictly regulated allocation of new cases between judges;
• the transparency of proceedings, including publishing not only motivated decisions itself but also dissenting opinions since reasoned decisions, as it is commonly believed, safeguard against judicial wrongdoing.

As for the transparency, we would like to explain its importance a bit more. Theodor Meron, who served as a judge for several international tribunals and as a President of some of it, underlined that both national and international judges for their independence should act in accordance with the transparency that belongs to the judicial process. By citing Justice Cullen, T. Meron stated that judicial arbitrariness is unlikely to have a place when nothing is hidden from the eyes and ears of laymen. Thus, the openness of proceedings not only helps a society to believe in a judicial system but also helps to restrain judges from unlawful actions which could devastate their reputation.

Finally, it should be mentioned that there are different approaches to accessing the level of judicial independency, while in this paper, we mentioned just the most popular and the most important criteria from our point of view. In our believe, these criteria are sufficient to understand the nature of judicial independence and to move to the major part of the essay, where we will try to imply and modify these provisions for AI courts.

4. Implementing Judicial Independence Criteria to Artificial Intelligence Systems

To begin with, from the first look it seems that being independent is an easier task for machines since they can straightly follow rules. Unlike human-judges, the work of AI systems can be influenced neither by changing their retirement age or salary nor by gifts or criminal charges. No one can informally call AI system and ask to decide a specific case in a specific manner. However, we believe that such risks should be mitigated in judicial AI systems as well, but with a correction due to its mechanical (as an opposite to human) nature.

We propose that implementation of the same criteria of independency as for human-courts with little corrections will enable to safe judicial independence even in the case of the use of AI systems in courts.

The first provision remains the same: both judicial independence and the judicial system structure should be fixed in the main legal acts. In the situation of using AI, this provision means that guarantees of judicial independence for AI managed cases should be stated in federal laws (e.g., procedural codes, judicial proceedings regulation, or special AI proceedings acts) and the application of AI systems shall be mentioned in major legal acts.

The second provision on the appointment of judges shall be transformed slightly. We propose to extend this provision not only on the development and the implementation of AI systems but also on the arrangement of the categories of cases to be resolved with the use of AI. All of these steps should be out of the direct control of one political or commercial actor. It means that, for example, the development of AI system shall be in the hands of several actors with different interests. Otherwise, we may get imbalanced systems for credit debt disputes, developed by organisations affiliated with banks; unfair systems for traffic fines disputes, developed by specialists affiliated with traffic police authorities; and, finally, pro-prosecution systems for criminal cases, developed without the involvement of attorneys. Similarly, it should be forbidden to empower one actor (e.g., the minister of justice or the supreme court) to decide for what cases and in which situations AI systems are applicable. 

Next, a fixed tenure shall be implied in this case too. The minimal period of using AI systems must be specified in advance, it should be sufficient (at least, three years), and it cannot be reduced without serious legal concerns (about this opportunity see the following paragraph below). If a legislative body decided to start a five-years experiment on adjudicating specific disputes with the use if an AI system, this term cannot be reduced later with concerns like economical inefficiency and so on (which can be only the visible part of real reasons). Such stability and confidence will assure AI systems’ operators and other involved actors that they will not lose their positions no matter which decisions their product will create.

The fourth provision, devoted to a removal procedure, is linked to the previous one. The list of grounds to dismiss AI systems from resolving the specific category of cases should be published in advance. This list should include only really harmful and unambiguous reasons, and the dismission of AI system should be in the arms of several actors with different interests at the time. However, higher courts should have an opportunity to produce interim measures in the form of the suspension of AI decisions if it can sufficiently violate human rights in a particular case.

Then, moving to the criteria of budgeting, we propose to extend this provision. Firstly, the salaries of personnel involved in the maintenance of judicial AI systems shall not be reduced and shall be increased the same way as the salaries of other similar specialists in courts (for judges helping to maintain AI systems – similar to other judges, for technical specialists and clerks – similar to other technical specialists and clerks, correspondingly). Unlike ordinary judicial proceedings, in AI managed proceedings, technical specialists play a much more important role. Technical job here is much more associated with corruption risks than in ordinary proceedings, where decisions rely only on judges. This is especially true when it comes to complicated AI systems which resolve disputes itself – in that case judges just help IT-specialists to design the proper model of resolving cases, while the operation of system in a particular case is more dependent on clerks and IT-specialists. Secondly, the budget dedicated to AI systems’ maintenance should stay sufficient, so it cannot be reduced during the same stated in advance period, and, more importantly, it should be increased due to the growth of the load of a system or to the growth of electronic component prices.

 The sixth provision is probably the easiest one. There should be a detailed regulation on allocation cases between human-judges and AI systems. No discretion should be left in this matter. All cases, which are fully consistent with criteria for being resolved by AI systems, shall be resolved by these systems.

Finally, we come to the transparency of proceedings. For AI-managed proceedings this provision means that (1) it should be clearly disclosed in the beginning of a case to the parties and in the decision on a case that AI system is involved and should be described how (in what part) it is involved; (2) the logic of an AI system’s decision should be described as far as it is technically possible. The latter highly depends on characteristics of a corresponding system, but where the transparency is technically possible, it should be provided. For example, if an AI system decided that 20% fine is inadequate for a 5-day delay in delivering package, the system, in the ideal world, should share the citations of judicial acts or legal acts, from where this opinion has arrived. The same way as reasoned decisions help to protect judges from wrongdoing, the necessity of providing explanations for AI decisions will safeguard the system from just typing in preferred results. Nevertheless, AI systems with unreasoned decisions may also be allowed, but the code of such systems should be a subject to a multi-actor control. In other words, if not in the decisions of AI systems, then the transparency should be reached in the code of the systems.

To sum up, all provisions, which were originally created to protect judicial independence in ordinary proceedings, should be applied to dispute resolution with the use of AI systems, but after various corrections of these provisions.

However, the list of criteria mentioned above would be incomplete without safeguarding the independence of AI systems itself. Previously, we indicated provisions, which can guarantee the independence of involved judges, technical specialists, budgets, and so on, but we still have not protected the systems itself.

Consequently, we propose to use nine criteria for ensuring judicial independence in AI-assisted proceedings. The eighth special criterion we propose for such situations is the protection of information from secret intervention from technical specialists, judicial clerks, and judges. This criterion means that nobody should have access to silently modify the results of AI systems work or to anonymously modify the inputs of a case in the system. We highlight here that mentioned criterion prohibits only situations when human-modified data pretend to look like AI-generated data or like the data generated by the parties to a case that should be used by AI systems.

Meanwhile, some of AI systems may require human assistance, which can lead, for example, to changes in AI-created decisions, especially on the first stages of using such systems. The most important point here is that such changes should be transparently demonstrated: who, why, and how exactly modified the work of AI systems. The same is for the modification of inputs: clerks should not have the ability to secretly change the data which was intended to be used by AI systems. However, the data can be reasonably modified: for example, if the system was mistaken that in the scan of a party’s document said “sent”, but a clerk clearly see that this word is “sell”, they can correct it, and everyone should see: who, when, why and what changed in the input data. In both cases, such intervention should be subject to judicial control when the party to a case is disagree with changes.

The ninth criterion we propose for AI systems’ independence is the protection from one-actor changes into the scheme of resolving disputes. This rule means that neither on the stage of creating AI systems nor on the stage of practical using it individual actors shall have the opportunity to directly influence the way of deciding cases. And it goes without saying that such changes in any case can be made only within judicial system or with a help of the council of legal professionals. Like the procedure of the appointment of judges by the executive branch of a government may be inconsistent with judicial independence, the same way creating and maintaining AI systems by the executive can endanger the independence of the judiciary.

Undoubtedly, individual judges can influence the practice via creating precedents, but it should be prohibited for individual judges and other specialists to simply modify the scheme of the work of AI systems. There is only one exception from this provision: when judges are responsible for the decisions of AI systems, and they shall check and correct all AI systems’ mistakes. Only in this situation, individual judges may have opportunity to fix legal mistakes (as they think) for all further cases of the judge for whom AI will provide an assistance. However, it is inadmissible for some individual senior judge to make such changes into the system in general.

All in all, we believe that these nine criteria can safeguard judicial independence in proceedings, where AI is used in order to help judges or to resolve some categories of cases instead of judges. Finally, we will approach the definition of judicial independence, which we provided in this paper earlier. In our opinion, AI systems consistent with these nine criteria are admissible from the point of view of the definition of judicial independence. AI systems, as well as judges who get the assistance from AI systems, will be free from external pressure and will be able to decide cases in the course of law without any harmful influence on them if these nine criteria are met.

Conclusion

Judicial independence plays an essential role in modern proceedings. The use of AI systems will lead to increasing difficulty in securing the independence of a judicial system because not only one entity, judges, needs protection from external influence, but also technical specialists, judicial clerks, and even AI systems themselves. However, this goal can be attained by adhering to the aforementioned nine conditions for ensuring judicial independence. Provided that the nine criteria outlined are followed, the independence of a judicial system should be ensured even when AI systems are employed.

Д.О. Дрозд, адвокат, тәуелсіз зерттеуші (Измир қаласы, Түркия): Судьялардың тәуелсіздігі және соттарда жасанды интеллект пайдалану.

Осы зерттеу тақырыбының өзектілігі сот төрелігін іске асыру саласында компьютерлік технологияларды пайдалану мәселесін ғылыми тұрғыдан пысықтаудың жеткіліксіздігінен туындайды. Жасанды интеллект жүйелерінің (бұдан әрі  ЖИ) дамуымен олардың сот ісін жүргізу қағидаттарын сақтауға ықтимал әсерін зерттеу қажеттілігі де өсуде. Әртүрлі елдер дауларды шешу саласына ЖИ технологияларын енгізуге тырысады, бұл процестер қарқынды жүріп жатыр, алайда бұл тақырып бойынша заң әдебиеті бұрынғысынша елеулі олқылықтарды қамтиды.

Осы жұмыстың мәні сот қызметінде ЖИ жүйелерін қолданудың сот тәуелсіздігі қағидатын іске асыруға әсерін білдіреді. Жұмыстың мақсаты ЖИ технологияларының судьялардың тәуелсіздігіне әсерін зерттеу және белгілі бір өлшемдерді сақтау сандық дәуірде сот билігінің дербестігін қолдауға қалай көмектесетінін зерттеу болып табылады.

Аталған зерттеудің тақырыбы қаралып отырған проблематикаға арналған заң жұмыстарының жоқтығының арқасында жоғары жаңалығымен ерекшеленеді. Әзірге көптеген зерттеушілер судьялардың тәуелсіздігі тақырыбына шоғырланып, оны құқықтық тәртіпті сақтауда тұрған негізгі қағидалардың бірі ретінде әділ ұстанады, сонымен қатар ғалымдардың көбеюі соттарда ЖИ технологияларын қолдану мәселесін әзірлеуде. Алайда бұл ғылыми еңбектер көбінесе соттарда және басқа да жалпы мәселелерде ЖИ жүйелерін қолданудың ықтимал ықтимал тәсілдеріне шоғырланады. Ал тек біраз жұмыстар ғана соттарда ЖИ қолдануға байланысты заңдық тәуекелдерді талдауға арналған. Бұл мақала ЖИ элементтерін қолдана отырып, сот ісін жүргізудегі судьялардың тәуелсіздігі тақырыбындағы келесі пікірталастарға есік ашатын алғашқы ғылыми еңбек болып табылады.

Осы жұмыста пайдаланылған зерттеу әдістері заңдық зерттеулер (талдау, синтез және т.б.) үшін типтік болып табылады. ЖИ технологиялары мен судьялардың тәуелсіздігі арасындағы қарым-қатынасты талдау арқылы бұл зерттеу сот өндірісіне осындай жүйелерді енгізумен байланысты болатын қиындықтарға жарық түсіруге тырысады. Осы зерттеу судьялардың тәуелсіздігінің түйінді сипаттамаларын анықтауға әртүрлі тәсілдерге жүгінеді, содан кейін оларды сақтау ЖИ жүйесін енгізу кезінде сот билігінің тәуелсіздігін қамтамасыз етуге мүмкіндік беретін мамандандырылған критерийлердің тізбесін ұсынады.

Бұл зерттеудің негізгі қорытындылары судьялардың тәуелсіздігін көрсететін критерийлердің мазмұны мәселесі бойынша доктриналдық пікірталасқа үлес қосады, сондай-ақ ЖИ-ті пайдалана отырып, сот процесі үшін арнайы екі жаңа авторлық критерий енгізеді. Автор тоғыз өлшемнен тұратын осы тізбені ұстану, атап айтқанда, сот билігі жүйесінің тәуелсіздігін қамтамасыз етуге қабілетті деп болжайды. ЖИ технологияларын енгізу тұрғысында судьялардың тәуелсіздігін сақтау критерийлерін қарау нәтижесінде осы бап заң шығарушылар мен өзге де жауапты тұлғалар үшін бағалы қорытындылар мен ұсынымдарды ұсыну мақсатын көздейді.

Кілт сөздер: жасанды интеллект; судьялардың тәуелсіздігі; соттардағы ЖИ; цифрлық судья; соттардың тәуелсіздігі.

Д.О. Дрозд, адвокат, независимый исследователь (г. Измир, Турция): Независимость судей и использование искусственного интеллекта в судах.

Актуальность темы настоящего исследования исходит из недостаточности научной проработки вопроса использования компьютерных технологий в сфере отправления правосудия. С развитием систем искусственного интеллекта (далее – ИИ) растет и потребность исследования их возможного влияния на соблюдение принципов судопроизводства. Разные страны предпринимают попытки внедрения технологий ИИ в сферу разрешения споров, эти процессы происходят стремительно, однако юридическая литература по данной теме по-прежнему содержит существенные пробелы.

Предмет настоящей работы представляет собой влияние применения систем ИИ в судебной деятельности на реализацию принципа независимости суда. Цель работы заключается в исследовании воздействия технологий ИИ на независимость судей и изучении того, как соблюдение определенных критериев может помочь поддерживать автономность судебной власти в цифровую эпоху.

Тема данного исследования отличается высокой новизной благодаря отсутствию юридических работ, посвященных рассматриваемой проблематике. Пока множество исследователей сосредоточены на теме независимости судей, справедливо позиционируя её в качестве одного из ключевых принципов, стоящих на страже правопорядка, в то же время постоянно растущее количество ученых разрабатывают вопрос применения технологий ИИ в судах. Однако эти научные труды преимущественно концентрируются на потенциально возможных способах применения систем ИИ в судах и на других общих вопросах. Тогда как лишь совсем немногие работы посвящены анализу юридических рисков, связанных с применением ИИ в судах. Данная статья является первым научным трудом, который открывает дверь для последующих дискуссий на тему независимости судей в судопроизводстве с применением элементов ИИ.

Методы исследования, использованные в настоящей работе, являются типовыми для юридических исследований (анализ, синтез и др.). Посредством анализа отношения между технологиями ИИ и независимостью судей данное исследование пытается пролить свет на те сложности, которые будут связаны с внедрением подобных систем в судопроизводство. Настоящее исследование обращается к разным подходам к определению ключевых характеристик независимости судей, а затем предлагает перечень специализированных критериев, соблюдение которых позволит обеспечить независимость судебной власти при внедрении систем ИИ.

Основные выводы этого исследования вносят вклад в доктринальную дискуссию по вопросу содержания критериев, отражающих независимость судей, а также привносит два новых авторских критерия специально для судебного процесса с использованием ИИ. Автор предполагает, что следование этому перечню из девяти критериев способно обеспечить независимость систем ИИ в частности и судебной власти в целом. В результате рассмотрения критериев сохранения независимости судей в контексте внедрения технологий ИИ данная статья преследует целью предложить ценные выводы и рекомендации для законодателей и иных ответственных лиц.

Ключевые слова: искусственный интеллект; независимость судей; ИИ в судах; цифровой судья; независимость судов.

References:

• Burbank S.B. Judicial Independence, Judicial Accountability & Interbranch Relations // Daedalus. 2008. No. 137 (4). – P. 16-27.
• Clark T.S. The limits of judicial independence. Cambridge University Press. 2010. – 337 p.
• de Oliveira L.F., da Silva Gomes A., Enes Y., Castelo Branco T.V., Pires R.P., Bolzon A., Demo G. Path and future of artificial intelligence in the field of justice: A systematic literature review and a research agenda // SN Social Sciences. 2022. No. 2 (9). – P. 1-21.
• Dunoff J.L., Pollack M.A. The Judicial Trilemma // The American Journal of International Law. 2017. No. 111 (2). – P. 225-276.
• Ervin S.J. Separation of Powers: Judicial Independence // Law and Contemporary Problems. 1970. No. 35 (1). – P. 108-127.
• Feld L.P., Voigt S. Economic growth and judicial independence: cross-country evidence using a new set of indicators // European Journal of Political Economy, 2003. No. 19 (3). – P. 497-527.
• Gewirtz P. Independence and accountability of courts // China Law Review. 2005. No. 1 (1). – P. 11-26.
• Gilbert M.D. Judicial Independence and Social Welfare // Michigan Law Review. 2014. No. 112 (4). – P. 575-625.
• Hamilton A. The judiciary department / The Federalist. 1788. No. 78. – P. 401-408.
• Hilbink L. The origins of positive judicial independence // World Politics. 2012. No. 64 (4). – P. 587-621.
• Kaufman I.R. Chilling Judicial Independence // The Yale Law Journal. 1979. No. 88 (4). – P. 681-716.
• Kaufman I.R. The essence of judicial independence // Columbia Law Review. 1980. No. 80 (4). – P. 671-701.
• Melton J., Ginsburg T. Does De Jure Judicial Independence Really Matter? A Reevaluation of Explanations for Judicial Independence // Journal of Law and Courts. 2014. No. 2 (2). – P. 187-217.
• Meron T. Judicial Independence and Impartiality in International Criminal Tribunals // The American Journal of International Law. 2005. No. 99 (2). – P. 359-369.
• Posner E.A., Yoo J.C. Judicial Independence in International Tribunals // California Law Review, 2005. No. 93 (1). – P. 1-74.
• Randazzo K.A., Gibier D.M., Reid R. Examining the Development of Judicial Independence // Political Research Quarterly. 2016. No. 69 (3). – P. 583-593.
• Ríos-Figueroa J., Staton J.K. An Evaluation of Cross-National Measures of Judicial Independence // The Journal of Law, Economics, and Organization. 2014. Vol. 30, Iss. 1. – P. 104-137.
• Shapiro M. Judicial Independence: New Challenges in Established Nations // Indiana Journal of Global Legal Studies. 2013. No. 20 (1). – P. 253-277.