Artificial intelligence has become an important turning point of technological developments. Especially with the use of this technology a transformation process has begun in our lives. Systems developed and under development together with artificial intelligence technology: for example, the development of autonomous systems narrows the areas of use of human power is spreading rapidly to societies and causing institutions to undergo changes on a broader scale. In this regard, it will be important to make technical and political judgements in the use of artificial intelligence as a possible governance model and to examine what kind of activities will have artificial intelligence-based standards in the current circumstances and how they will contribute to artificial intelligence governance on a global scale. Indeed transformations experienced in the economic and social context have also affected the building blocks of the international order. In this sense, transformations such as a global governance without a state, without borders, new ways of life have emerged and for human activities in sectors such as morality, law, business, education and health also involved in these transformations. Besides such issues it is important to determine the guidelines to be followed in the transition from the current order of the nation-state actor in the political dimension so-called the modern period to the stage called the post-modern period and to determine what the principles of post-modern security are. In this context, along with the transformative development of artificial intelligence technology, it is necessary to take into account such features as the ability to act as a mediator in the dispute resolution of international agenda such abilities are e.g. ability to develop autonomous war machines to the extent that they can change the form of wars. In addition, the fact that the digital infrastructures that many of the states have adapted to, namely cyberspace, have also become a new dimention of security and that reveals the importance of the idea of an artificial intelligence-based governance model.

Keywords: Artifical Intelligence, International Security, Artifical Intelligence Governance

Introduction

The main question in this paper is whether an international or supranational governance model based on artificial intelligence technology can be achieved or not. In this regard, first of all, the concepts of artificial intelligence, governance and autonomous will be explained in order to examine the phenomenon to explain much clear possibility of an artificial intelligence-based governance mechanism for the settlement of international relations disputes. Then, the current status of international standards of artificial intelligence systems will be briefly mentioned. In this way, based on the question of what are the governance mechanisms; first of all, the concepts of the state, state control devices, constitutionalism and artificial intelligence constitutionalism will be clarified. In the next stages, the concept of international security and the effects of artificial intelligence on international security will be explained. Finally, it will be discussed what the artificial intelligence governance model can be like or whether there is an agreed conclusion that can be reached within the possibilities.

Artifical Intelligence and Governance

Artificial intelligence is system that may perform as human intelligence for particular tasks generally by repeating the information that have collected. Simply, the transaction process can be described in three important stages in outline. Artificial neural networks created in the stage of data processing and transferring information that is to create output by processing data and models in an analytical way are essentially part of the transition to the development of the concept of deep learning, which is the next stage of what called as deduction. Deep learning is in this regard, one of the important stages in which the qualifications of the output are determined in the process of making sense of artificial intelligence and learning. When we refer to difference between governance and government it is that the concept of government is a group of people or a cabinet that controls the mechanism that has the means of governance. However, the concept of governance is a phenomenon in which a number of rules and framework ideas are drawn. The concept of artificial intelligence governance in this regard is the whole of the ideas presented or the processes carried out in the government or management of artificial intelligence especially in the international scale and it is about what kind of governance is in there and which set of rules, will be put forward and how they can be applied. In this regard, artificial intelligence governance can be determined by automatic, automatic or autonomous types. To clarify these phenomenons: the term automatic artificial intelligence systems are the simplest mechanisms that have been developed. These can be exemplified from a toaster to a wire that allows the arm to be triggered. The concept of an automat can actually be an operational process or a vehicle, it is a process of control and execution, the simplest example of this is the speed control systems of vehicles. Autonomy, on the other hand, is to take precautions against unexpected situations. This can be a simple task, such as preventing the shutdown of any system, and the autonomous structure should be able to create spontaneous solutions by modeling the decision-making mechanism in the system independently of the controller (the human). In this context, in order to develop an artificial intelligence-based approach at disputes solving, autonomous artificial intelligence should be taken as a reference at this point.
Along with artificial intelligence, many technological developments have emerged. (Internet of Things or big data concepts) In this regard, the stages of development and use of artificial intelligence technology may differ depending on regions and countries but it has not yet fully reached a wide range of rules and standards in international specialization. However, in the diffusion of technology, there are standards that regulate artificial intelligence procedures, especially from a technical point of view, because the countries that provide new acquisitions ensure the adaptation of this technology by providing imitation. Organizations that provide such standards can be given as ISO, IEEE and CENELEC. However, use of it in international relations within scope of security and diplomacy, there are no common standards, but it is observed that each state carry out different strategies. For example, the People's Republic of China uses artificial intelligence as a digital assistant within the scope of its artificial intelligence strategy: it receives support in matters such as ensuring cyber security, threat impression and social control. It is certain that artificial intelligence is just beginning to acquire influence for pervading most domain of international relations.

Governance Mechanisms

In essence, the concept of governance processes decision-making processes within the action and ability to manage or dominate a particular organization. While providing the concept of an artificial intelligence governance, first of all, the study of the state, the main actor of international relations involved in this governance process, will shed light on the issue. Defining the state which i refer to modern state, is fact formed after the Peace of Westphalia. The main thing here was to ensure border control, that is, the state apparatus was creating an embodiment by ensuring territorial dominance, drawing its power with borders and creating a physical reality for people. The state also created the central authority and held the monopoly on violence. When people tacitly conceded their security rights to the state, they essentially created the Leviathan. A number of institutions have been developed to consolidate the power and ensure control in the structure in which the state is organized in this way. In addition to the presence of security forces inside and outside as a deterrent force to prevent conflict and sustain the power, the state has consolidated its control by creating a bureaucracy and establishing a mechanism that extends to the social sphere. Other important the characteristics of statehood evolved in 19th and 20th may be incorporated into as: rule of law and forming public sphere.

In fact, it will be laws that now to examine in this regard: By systematically writing commands and restrictions into law, states have realized their legitimacy in a constitutional dimension, unlike the linking to divinity or references to the sanctity of the seigneur seen in previous systems. In addition, constitutionalism, observance of laws and subordination, which are often seen especially in Westphalian state models, should come to the fore; this concept, which is not seen in every state model but is also provided by case law with unwritten tradition, will be important to understand whether states should also commit to standards or norms in international relations. As such, although constitutionalism is associated with democracy with its current use, it goes back further in modern democracies, and constitutionalism exerts a wider range of influences than democracy implies. In this sense, the constitution is “the law for the laws“; it is the highest of the laws, the source of the laws. There are elements such as the rule of law, the constitutional human rights regime. Then when referring to Constitutional government, it is the opposite of the unlimited arbitrary use of authority and the performance of the administrative function by instructions, not by rules. The essence of the phenomenon is that in order for the power of the state not to lead to domination in the majority, the power of the state power has been restricted on a constitutional basis of individual and social freedoms. In the formation of an artificial intelligence-based governance similar to constitutionalism, in the creation of international laws; creating an analogy about the fact that countries are subject to a constitution conducted with artificial intelligence reasoning will be an important situation that can be examined for the issues that will be discussed later. First of all, it is necessary to define the phenomenon of an artificial intelligence-based constitution. This term will simply cover that artificial intelligence technology will make to setting norms and maintaining and maintaining existing judgments. In this regard, artifical intelligence based constitutionalism will cover human value judgments and social principles in essence. While determining the constitution, the processes that are in the preparation of a constitution can be carried out in the same way. For example, for any constitution, it can be clearly stated which behaviors of people will be rewarded and which will be punished. In this respect, constitutionalism can be one of the building blocks in the development of governance models. Because the concept of artifical intelligence based constitutionalism is a phenomenon developed in order to ensure the governance of this technology. If the constitution that can be designed should be fundamentally democratic, it should also be open to negotiation at a level that can be subject to change. This should also be integrated into every stage of social systems. Therefore, the purpose of using artifical intelligence technology should be created in such a way as to provide social and economic benefits to people. In this regard, governments should prioritize human benefit and security over their other priorities such as industrial development. Because the development of a misprojected set of laws in this direction may lead to events that may negatively affect society, such as mass unemployment, and may lead to non-state oligopoly utilitarian groups that hold technological power for the direct government decisions to increase their influence on the government. In fact, when it comes to any issue, the scope of the phenomena develops by simply evolving from the micro level to the macro level.

In this regard, if it is desired to ensure governance in an international context, human activities are at the basis of the formation of complex structures from the constitution to the state administration.

International Security and Artifical Intelligence

Defining the concept of security, it is that absence of a state of danger or fear, or potentially the extinction of a state of danger. In fact, the concept of security has traditionally been interpreted: security forces ensure public order within the borders of the sovereign state and protect its territorial integrity, sovereign rights against external threats. Another dimension of the security issue is that the concept of security has gained different dimensions by breaking away from the traditional framework with the free market economy that developed after the Cold War and the rise of the global economy. In the current literature, Realist theory, i.e. conflictual environment, and neoliberal theory, i.e. cooperation, can be mentioned mainly in the theories developed on ensuring international security and peace. While realists explain the international security environment with the concept of anarchy, there are claims that states maintain their confrontational identity by following their national interests only. In perspective of neoliberalism there is a belief that cooperation and cooperation are possible where states are not constantly in conflict, but there will be no losers. Here in fact, the belief that international security can evolve into a peaceful environment will create an environment of cooperation, since the increase in the level of dependence of states or actors, especially their economic relations with each other, will make each actor fragile to negative outcomes. As well as these interpretation of environment of security. There are some periodical stages of security term. If the process of evolution of the concept is categorized in general: it can be divided into traditional (Pre-modern), Modern and Post-Modern periods. In the first of these, in the traditional societies the roles (traditions) given for generations are lived without question, there is no distinction between a clergyman and a scientist and people are far from creating a utopia. O the other hand, modern stage is the period that is assumed to have started with the emergence of the idea of the nation-state. Modernism: it includes industrial economy, capitalism and class systems. Currently it may be called as post-modern period. This is breaking down political and economic models at the macro level by reducing them and restructuring them by providing global system integration. It is a period where conflicts no longer take place completely and at the border area under the domain of a nation-state actor, as in the modern world, and armies are becoming increasingly marginal. In addition to these changes that can be observed in the army structure, the control network of the modern world has also been shifted to intelligence systems. While intelligence is coming up differently then army, it is evaluated at two different points. In the same way, the formation of a multi-environment of actors not only states, but also supranational, non-state actors, multinational companies, makes it difficult to determine the positions of actors in the existing intelligence network. One of the critical elements of this transformation of the post-modern period is the emergence of a technology called as “Internet“ in the process from innovations such as IT technology, microprocessors, developed especially in the 1970s and 1980s. In this regard, the commercial, social and political changes created by the industrial revolution will subject to changes in the post-modern period, especially in international relations. Thus artifical intelligence can contribute in areas such as public diplomacy, resolution of inter-states disputes, policy production processes. In this regard, in the international security environment of artificial intelligence technology; transformations can also be achieved in ensuring security.

In this regard, assuming that AI governance is an organization in international context then the functions of this organization can be briefly mentioned. First of all, we can divide an organization into three basic functions. Firstly, it is “Obligations“ the function of indicating the responsibilities of actors determined in accordance with the scope and goals of the organization. Secondly, “Compliance“ it is measuring in what scale commitment brought by actors, demonstrate to the obligations imposed by the organization. Thirdly, “Enforcement“ it is the creation and implementation of sanctions against states that violate or fail to implement obligations. In this regard, the “Online Dispute Resolution System“ system can be inspiring in developing an idea of how successful artificial intelligence technology can be in dispute resolution in an organizational dimension. [9]

As can be seen in this template, the negotiations within the system continue until a solution is reached. If the negotiations are inconclusive, the issue will be resolved by mediation. However, there are five main problems that ODR systems can face. As follows: (1) Recognition of the problem (2) Understanding of preferences (3) Effective communication (4) Ensuring neutrality (5) Degree of Autonomy.

Indeed although such systems are useful in diagnosing problems in many sectoral, national and international areas and reaching alternative solutions, they have not yet reached a level of autonomy. If it is necessary to create an analogy to the issue in the context of international security, disputes may occur between different actors. But generally assuming that the state actor is at the forefront. We can take an example of dispute. For example, let's examine the case of the Aegean Sea between Turkey and Greece. Greece wants to expand its maritime borders to 12 miles and Turkey is against this issue and let's examine with a thought experiment how artificial intelligence autonomy and neutrality can have an impact on disputes. First of all, the problem needs to be diagnosed. Many of the disputes between the parties may be a misinterpretation of intentions, which can also be explained by the concept of security dilemma. In this example, Greece may want to use a legal right that exists in accordance with the UNCLOS maritime convention, but when asked why it wants to use this right against the other side and not in a way that protects the status quo, Turkey may attribute the issue to Greece having an idea of territorial expansionism. It is possible for the parties to diagnose a common problem in this regard, but they have problems explaining their preferences clearly, and the fact that effective communication channels are closed especially internal and external political fluctuations or economic and social developments can be effective in this case. Since the parties are into the negotiation, there is also a problem with the principle of neutrality. Whether it is considered as a decision of the body that will decide the issue, for example, the UN International court of justice, the parties will perceive the degree of autonomy of the mediation institution differently in this case. In fact the parties that are the subject of the issue are not pure metas but state or governments: hierarchly over formed by human activities that have borders, have geopolitical strategies, are managed by government cabinets, have characteristic feature which represents its society, norms and values. In essence, complex entities at all levels. In this regard, the trust and benefit that a human institution will offer will be limited. In particular, since our data processing capacity can not be as high as the computers we have developed, it will not be possible to consciously perceive such data simultaneously and to reason by keeping all this in our brain memories. This is where the belief that such technologies can provide efficiency in dispute resolution is born. If this computer technology, which has an incredible data processing and storage capacity exceeds the threshold of sub-parameters that should be taken into account in resolving or facilitating the resolution of a dispute, this will already give the system a certain degree of autonomy. The technology that will gain autonomy is not only computer technology that performs calculations but also what we currently call artificial intelligence with progressive algorithms based on data processing and imitation; it is a technology that contains subsets such as deep learning.

Capabilites of Artificial Intelligence

First of all, it is necessary to reveal the assumptions about the direction of the developments in the technical capacity of artificial intelligence. In fact, there will be two main questions to be asked: (1) speed of development in artificial intelligence technology and (2) what kind of research fields it can expand into. As for what kind of research areas artificial intelligence can access, there are two main questions: (1) Will artificial intelligence develop as a superintelligence model or (2) will it follow incrementalist dimensions. In this regard, the features that can be expected from an artificial intelligence that can have a technology with super human intelligence will be transformative in essence. To give an example of the areas where such transformative effects can occur: cyber security, the autonomous weapons industry, finance, production, science and technology can be explained in the form of. But on the contrary, the impact of incrementalist artificial intelligence technology on such fields’ impact will have will largely remain at the level of debugging and performance enhancement. There is a reality underlying the prediction of such a narrow-scale artificial intelligence development. For example given the fragmentation of the artificial intelligence industry today and the uneven diffusion of its development in different contexts, this kind of argument is actually a more realistic one. The most concrete example to be presented about this argument is that many artificial intelligence technologies have the capabilities of being able to translate, such as driving a car, but still fail in experiments such as the Wingrad Scheme. In addition to technical capabilities, another issue that should be mentioned about ensuring the reliability of artificial intelligence systems will be artificial intelligence security. If it is to be defined, this concept refers to the purpose of moving artificial intelligence technology to a safe and useful advanced level. Just as an engineer wants to make sure that a bridge he is building will not collapse, such a safety control mechanism is also needed in artificial intelligence systems. For example, there is a risk of accidents and unexpected behaviors in simple artificial intelligence systems and exclusive and complex systems should be developed to prevent them. In this regard, the basic features of a mechanism that can be established to ensure the security of artificial intelligence efficiently can be listed as follows:

Conclusion

If artificial intelligence technology is reinterpreted in the international security dimension, the realist theory perspective will offer a direct contribution to the impact of artificial intelligence on international security. The main reason for this is that in the realist theory of the confrontational environment, modern states all have professional security forces and are seeking to develop weapons at a level that will increase deterrence and lethality in this direction. In this regard, artificial intelligence technology, especially from the development of autonomous autonomous aerial vehicles and land vehicles, features such as detecting enemy threats with deep learning and attacking targets with autonomous initiative, will transform the quantitative and qualitative calculations of states when using security forces and warfare. It is necessary to give an example in this regard, if a state experiencing a border conflict with its neighboring country has such a technology, if one of this cpuntry has autonomous attack forces without loss of life, or advanced air defense systems will also ensure that this country is vigilant against possible attacks, in fact, this technology has a tendency to lead to a new arms race that will strengthen the state's sense of security. In this regard, they have shown the necessary importance of this technology in the states and have determined missions and visions by publishing national artificial intelligence strategies. In contrast, this technology can be an mediator in the issues regarding “security dilemma“ in the case of a bilateral autonomous cyber technology that can be established between actors, the parties can provide an environment in which data flows between each other can be clearly expressed. Because, intelligence or security intelligence activities as a means of collecting information have caused states to fall into a security paranoia since they have created a system in which states follow each other one after another. In this regard, if there is an international institution to ensure artificial intelligence governance, first of all, this institution should based on cooperation to be used by encouraging between states, and not as a tool of domination and forcing. However, determining the degree to which governance will be centralized will be another situation that will determine. Because concepts such as the “decentralized center“ contained in the blockchain technology, which is a subset of artificial intelligence technology will be important in determining the functioning of the institution. In this regard, it would be the well if the governance mechanism that needs to be established takes on a central role. Of course, the issue that should be stated about whether a governance without a state or independent of actors is possible, which is the focus of their discussions, is this: if the concept of a decentralized center is developed, a consensus obtained will not provide an efficient method of resolving incompatibilities or improving the quality of decision-making processes. However, the central functioning mentioned here is different from the concept of monocenter execution of artificial intelligence technology. If we take into account today's realities, we know that this technology has not the same level for every actor, especially each state has access to this technology at different levels. Therefore, it is necessary to create standards of specific areas that will be recognized by the masses. Based on the example of constitutionalism: ensuring sectoral confidentiality, especially in areas such as finance and health, a system is needed in which national artificial intelligence can be used without malicous use by state actors and international data sovereignty can be achieved technological information flows independently from borders. According to Political scientist Gaetano Mosca, every socio-political regime has the potential to lead to the power of an elite minority, knowing that artificial intelligence technology has actors: states and especially private companies should ensure the distribution of activities in this technology with a wide-ranging governance. I would like to emphasize especially private companies here, while the state is considered the most important actor in the international arena, private organizations can come across as oligopolistic tyrants of this cyberspace. Especially companies like Facebook can tend to gain influence by consuming people's data almost as a monopoly, as in the case of “Cambridge Analytica“. Therefore, against the trend of hollowing out the concept of the state and see state as completely coercion tool; in fact, it should be realized that the state is one of the most comprehensive and functional social institutions created by mankind, and its rights and freedoms should also be realized by ensuring adaptation to artificial intelligence technology or cyberspace through the state apparatus.
In this respect, it will be necessary to build governance arrangements based on the following requirements in short.

References