The ex-ante regulation of artificial intelligence in Ethiopia

Master's Thesis, 2019
92 Pages, Grade: very good


Table of Contents

Acronyms and Abbreviations viii Abstract . ix

Chapter –One
1. The Introduction
1.1. Background
Table 1: Estimated share of employment that is susceptible to automation in key countries of interest
1.2. Statement of the Problem
1.3. Objectives of the Research
1.3.1. General Objectives
1.3.2. Specific Objectives
1.4. Research Questions
1.4.1. Main Question
1.4.2. Specific Questions
1.5. Significance of the Study
1.6. Research Method and Methodology
1.6.1. Research Methodology
1.6.2. Research Method
1.7. Limitation of the Study
1.8. Scope of the Study
1.9. Organization of the Chapters
1.10. Ethical Consideration of the Study

Chapter Two
2. Conceptual and Theoretical Framework of Artificial Intelligence
2.1. Definition and Nature of Artificial Intelligence
2.2. Philosophy and Theory of Artificial Intelligence
2.3. History of Artificial Intelligence
2.4. The Interplay between Science, Law, and Technology
2.5. Principles and Rule

Chapter Three
3. The Regulation of Artificial Intelligence
3.1. General Concepts of Regulation and ex-ante Regulation
3.1.1. Definition of Regulation
3.1.2. Definition of ex- ante Regulation
3.1.3. Rationales behind Regulation and ex-ante Regulation
3.1.4. Advantages and Disadvantages of Regulating Artificial Intelligence...23 Advantages of Regulating AI Disadvantages of Regulating AI
3.1.5. Significance of ex-ante Regulation of Artificial Intelligence
3.2. Competent Government Organs to Regulate AI
3.2.1. The Legislature
3.2.2. The Executive
3.2.3. Suitable Regulatory Institution
3.4. Legal Liability of Artificial Intelligence
3.4.4. Allocation of AI Liability (comparative perspective)
3.5. Ethics of Artificial Intelligence

The Assessment of Development and Legal Framework of Artificial Intelligence in Ethiopia
4.1. The Embracement of Ethiopia for Artificial Intelligence
4.2. Current Impact of Artificial Intelligence in Ethiopia
4.3. Existing Regulations and Artificial Intelligence
4.3.1. General Policy Framework
4.3.2. Subsidiary Legislation The Ownership or Property Rights of Artificial intelligence
4.3.2. Tangibility of Artificial Intelligence
4.4. Challenges Posed by AI on Existing Jobs
Picture -4: Blue-collar workers figure



First and foremost I would like to thank almighty God in the name of Jesus Christ for what he did and doing in my life. This work would not have been possible without some very special people by my side, My advisor Bisrat Teklu Hialemicaheal, My parents, my brothers and my sister who were always there to give me their unconditional support, and provide me with their vast wisdom, My hubby, who helped me during the hard times, and always made me smile. My friend, Tseday Gebrhana Who made it all more joyful, and all my friends and relatives who gave me strength and motivation to start this journey.

I would also like to thank the University, the Artificial Intelligence Directorate Director Ato, Ashenafi Bekele, the i-Cog company manager Hruy Tsegaye, and all experts from National Bank of Ethiopia, Ethiopian Electric utility, MiNT, and the Ethiopian statics agency who made this possible through their support. In addition, I would also thank all my peers who gave me some invaluable feedback on my work.

My little angel Christian this is for you and I love you.

List of Tables

Table 1: Estimated share of employment that is susceptible to automation in key countries of interest

Picture -4: Blue collar workers figure

Acronyms and Abbreviations

Abbildung in dieser Leseprobe nicht enthalten


Artificial intelligence is one of the newly emerging technologies which are creating new challenges to the existing laws and raising serious survival questions. Regarding the relation of artificial intelligence and the law, there are serious concerns on how the law regulates artificial intelligence as it is now being more difficult for traditional public regulatory bodies to control the development of AI. Some form of regulation is likely necessary to protect society from harm. Due to the power and complexity of this new emerging technology, Regulation can, indeed, be very impactful, but it also carries risks. This thesis tried to solve this difficulty and examined different issue to answer whether artificial intelligence should be regulated or not and, if so, which basic principles should be followed and who are the suitable organs to regulate it. It also demonstrated the intricacy of this newly emerging technology as it has its own positive and negative consequences on the life of the society. Since the AI technology is prevailing constantly from time to time and its involvement is increasing in every aspect of activities such as factories or hospitals, the particular research emphasized the need for the development of AI that reconciles with its Ethical, legal as well as social issues. To this end, the issue of ex-ante regulation was underscored as a necessity to handle the potential challenges of AI in a formal way.

Chapter –One

1. The Introduction

1.1. Background

Inequality is the background to the countless ills of our society. The confidence that technology is our vehicle for escaping the curse of dramatic inequality that plagues the world today is increasingly gaining wider acceptance.1 In a very real sense, computers have proven capable of far more physical and mental “human” functions than most people believed were possible. Nowadays there is such an increasing similarity between human beings and machines that in fact eventually needs legal recognition of computers as “persons”.2 Put it in a simple way, the word of ‘Artificial intelligence’ (AI) can be used to computer arrangements which are planned to make an exact copy of human cognitive functions.3

In undertaking decisions which are the form like “IF this situation exists THEN take this action” is considered as a kind of artificially intelligent system4. Unfortunately, there does not yet come into reality or use to be any widely acknowledged explanation of artificial intelligence, even among experts in the particular division of artificial intelligence study, much less a useful working definition that will serve as an input in regulations. However, John McCarthy, who is well-known as the father of artificial intelligence and the computer scientist and who initially created the term AI, did not limit intelligence in AI to the replication of human intelligence, but argued that machines could display other intelligence that involves "much more computing than people can do." He defined AI as the science and engineering of creating intelligent machines, in particular intelligent computer programs."5 From another point of view, AI can “operates autonomously, perceive their environment, persist over an elongated period of time, and adapt to change, and create the best-expected outcome." 6 Or it is theory and progress of computer systems intelligent to carry out responsibilities that normally need or depends on human intelligence.7 Substantially, today’s artificial Intelligence lacks general intelligence of human beings in many aspects.8

In its essence the AI is the capacity of a computer program9 to perform functions usually related with the capacity to apply knowledge and skills of human beings, for instance, learning, understanding, reasoning and interacting, decision making, in other words, to "normally do good, justified or acceptable thing at the right time" 10 These days there are quite surprising functions noticed from AI. So that, AI systems extend undertaking tasks such as filtering email spam, passing recommendation for people what things to purchase, providing legal aid on many things ranging from parking tickets to asylum applications, etc...11

The FDRE prime ministry Dr. Abby Ahmed made the following speech about AI in his recent public speech and passes his advice and concern on the new discipline of AI for readiness for coming changes.

"We will face challenges in the near future, maybe after five or ten years, by the newly emerging technology which is called the artificial intelligence or the fourth generation industrial revolution. The important point is not the money we collect from the internet service but the data. In the near future, there will be no use of Wi-Fi or internets under geographical limitation. It will be necessary to connect all over the countries so that no one should be limited from speaking, writing and doing something because of internet accesses scarcity, creating the artificial intelligence machine would be is a necessity to get data through communication. The machines can learn faster than human beings, for instance, up to 10 or 15 medical physicians work on the treatment of one patient and get to the final point, but the artificially intelligent machines can substitute those physicians and probably work better than the human physicians with full competence. In a similar vein, AI has also the same impacts on other sectors such as agriculture and industry. Interestingly enough the new fourth generation, industrial development can bring such a remarkable change than the old industrial revolution development that counted five centuries. These days the trend of competition among the different countries of the world seems to resemble the adoption or introduction of artificial intelligence among both the industrialized and the developed countries with those amazing AI machines. However, there are limitations on those machines as they are not as human beings, such as they lack emotional intelligence and for this reason we Ethiopians as we have a good profile of social interaction and different cultures of our people, we can take this as an advantage in redefining the use of AI technology in a manner more beneficial to the society …."12

An array of current developments has shed light on the growing amount of special care or consideration being paid by legislators and other government authority or public body in the area. In fact, we are in the initial period of global intelligence revolut ion. But the AI already spread in many features of our lives.13

“In a world where nothing is perfect, humanity has struggled to make society work in a more desirable, satisfactory or effective way. One of the tools used for that is a regulation or directive made by an authority, for the purpose of guiding the behavior of people and objects.14 Even though knowing when to regulate and how to do it is a very complex task. For that, it is primordial to evaluate whether regulation would be appropriate for the case. Then develop a regulation such that it can be stared as successful in addressing the issue at stake”.15

Currently, as it is witnessed in many instances, the challenges posed by AI are bolder than another time before. For the first time, creations of the human beings are being capable of learning on their own, of deceiving and overwhelming the capabilities of their creators, of developing a distinguishing quality or original nature and ability without further action of intervening, and of acting with autonomy. These peculiars are eventually attributed or an aspect of artificial intelligence that is unique to the discipline and is again the sources of many challenges that spillover out of it. Law requires to be updated in order to coup up with these edge cutting technologies; otherwise, they would be obsolete and lag behind to address the recently discovered technological breakthrough. To this end, it is essential to achieve clearly defined or identified rules that suit these novel characteristics of Artificial intelligence.16

As it is known, these days in every angle of the world, technology grows fast and the understanding of the technology becomes more complex. The scientific progress may be unpredictable and it may be difficult to control, but carrying out proper studies on this filed with the aim to maximize social benefits derived from science is necessary.17 Additionally, scientific and technological improvement develops at an alarming pace faster than law.18 And because of the inattentiveness of our rules, and the power and complexity of such technology, both law and society have to be changed and adaptation to a control system must be built, and the state must intervene in the sequence of action to make sure that everyone is heard and every right is guaranteed.19 Moreover, the regulation does not need to be a “necessary evil”, in the other direction; it can bring many benefits to the science of AI and society in general. This is because of the reason that unpleasant legislation might create challenges such as disabling the growth of automation.20 But instead of such regulation it is necessary to take into consideration of a beneficial legislation, like making this technology more acceptable and accessible, improving risk access and management, support or encourage the growth of the interdisciplinary discussion of this field, and keep deter unnecessary outcomes from happening and causing damage due to technical or philosophical omission of expected or required action.21 The global report on AI and Life in 2030, states that artificial intelligence has the potential to challenge any number of legal assumption, for such reason governments and policymakers across the world are beginning to struggle and take part with what AI means for law and policy and the essential technical and legal frameworks.22

On the other side, when we come to Ethiopia, which is on the verge of its economic development that moves in a particular direction is highly prone to utilize the inputs of the modern technological advancements in its fast-growing economy in many of its major economic sectors. For instance, if we take the 2016 world development report Ethiopia, the country is on the top rank with 84.9 % of the automatable share of employment unadjusted for adoption and 43.9 adjusted for adoption followed by India and Nigeria even above the average of OECD which is 57.0 as shown by the following table:

Table 1: Estimated share of employment that is susceptible to automation in key countries of interest. 23

Automatable Share ofEmployment

Abbildung in dieser Leseprobe nicht enthalten

Although the World Bank has adjusted these figures based on an ‘adoption time lag’, it is with the contention t hat these predictions should be taken with further caution.24

In this case, there is a need for the enactment of appropriate legislation for the regulation of artificial intelligence for governing its overall aspects early. These legislations are advantageous since as labor becomes a less significant factor in production as compared to owning intellectual capital, a majority of citizens may find the importance of their labor not enough to pay for a socially acceptable standard of living.25

The purpose of the present research is to verify the marked effect or influence of artificial intelligence on the traditional categories of law. The study assesses if these categories of law can be used to regulate AI whether they are suitable or not. Finally, the researcher will attempt to propose legal solutions that may play an important role before the introduction and wide diffusion of such emerging technology, the challenges and the profits of its early regulation specifically concerning with its benefits and risk.

1.2. Statement of the Problem

There is a high probability that in the near term AI will replace tasks rather than jobs in the near future, and will result in the creation of new kinds of jobs. However, those new jobs that will appear are so difficult to imagine in advance than the existing jobs that will likely be lost or the distribution of the economic fruits of artificial intelligence impacts on the working force.26 Innovations usually improve safety and living conditions, therefore people do not agree to take the uncertainty that could affect their certainty.27

All over the places where the industry is founded, there are clear examples of industries in which digital AI technologies have had profound impacts, good and bad, and other sectors in which automation will likely make major variations in the near future. Many of these changes have been driven powerfully by “routine” digital AI technologies, containing enterprise resource arrangement, networking, information processing, and search. Understanding these modifications should afford visions into how AI will touch future labor demand, including the change in skill demands. Nowadays, the technologies in the digital world have been affecting workers, especially those in the skilled middle.28 So, this research in addition to the future labor request problems, it will also try to make a focus on such problematic issues that are in relation to data and external world interrelation. States such as China, Syria, and Iran are the only countries that rank worse in internet freedom than Ethiopia.29

As stated above, one of the unique features of AI that makes it different from previous technologies is its capability to act autonomously. Already, AI systems can perform complex tasks, such as driving a car and building an investment portfolio, without active human control or even supervision. Just as the Industrial Revolution caused socio - economic sudden change as mechanization reduced the necessity for human manual labor in manufacturing and agriculture, AI and related technological advances will decrease the demand for human labor. AI will force comparably disruptive changes to the law as the legal system struggles to cope up with the increasing ambiguity of autonomous machines. 30 The fact that AI also presents a systemic risk, which is the embedded danger to human health and the environment in a larger context of social, financial and economic risks and opportunities’, which requires the organization of numerous stakeholders in the monitoring process.31

In this case, the problem that arises is whether absence/lack of ex-ante regulation of AI impacts on creating a stable future of the society’s economic, political and social development. The potential risks that the AI results on the major problems stated above and the deficiency of adequate legal framework, the thesis suggest, the early preparation of regulation of artificial intelligence.32

1.3. Objectives of the Research

1.3.1. General Objectives

The general objective of this research paper is to discuss the status quo of artificial intelligence, regulation in Ethiopia with the advantages of ex-ante legislation on the matter of AI and it questions the proactive legal protection of the AI with its advantages and risks.

1.3.2. Specific Objectives

1. Discussing the advantages of regulating artificial intelligence and that of legislation on AI on an ex-ante basis.
2. The focus of this study will be on the influence of AI on people living in Ethiopia for whom many of the same opportunities and risks apply and examining the existing legal framework to perfectly address it.
3. The research aimed to disclose the challenges posed on certain kinds of professions as the reality of legal gaps can lead to problems which current law is not able to clarify, the liability issues raised when a smart machine causes physical damage to someone.
4. It also looks at the available remedies for human power replacement with Machines and the destruction followed this will be discussed.

1.4. Research Questions

1.4.1. Main Question

The main question of the research is what is the underlying rationale for regulating AI and The ex-ante regulation of Artificial intelligence in Ethiopia?

1.4.2. Specific Questions

This thesis attempt to answer the following question;

1. Are the existing legal frameworks capable or perfectly suited to address the subject matter of AI?
2. How regulating AI benefits? What are the possible rationales and challenges in regulating Artificial intelligence?
3. What are the remedies for human power replacement with Machines and the destruction followed?
4. How do questions of ethics in AI be reconciled that overlap with concerns about biases?

1.5. Significance of the Study

In every angle of the world technology grows fast and the understanding of the technology, even becomes complex for such reason governments and policymakers across the world are starting to struggle and deal with what AI means for law and policy and the necessary technical and legal frameworks33 as it is readily and widely known in Ethiopia there is a tremendous rate of unemployment at the national and regional level. It is also known that among the different factors that escalate this social crisis of unemployment is the substitution of ordinary labor by automation and other technological inputs. If we consider the 2016 world development report of the world the estimated portion of employment that is vulnerable to automation in Ethiopia is 84%, which is high as compared with the rest of countries in the top list like India& Nigeria and even greater than the OECD average which is 57%. Apart from this, the risks posed by uncontrolled utilization of AI are also expressed in its impact of weakening the usefulness of digital market.34

As a result, the findings of this research would contribute in reminding the legislature to give due attention for the promulgation of relevant AI national policy, on the foundation of which subsidiary laws would also be enacted. The study tries to disclose the main hurdles that are posed by the absence of proactive legislation in the field of AI in Ethiopia. In addition, it also tries to indicate that adequate attention should be given to infrastructural development.

1.6. Research Method and Methodology

1.6.1. Research Methodology

From the viewpoints of the objectives of the study, as legal research, it can be labeled as an exploratory, descriptive, experimental and doctrinal legal type of research.

1.6.2. Research Method

In this study interview and literature review were used as a research tool to gather data, the qualitative and quantitative Research were the methods in the study. The primary, secondary and tertiary sources are also a source of information in the research; the key tools, such as Literature review, Laws, other document analysis, interview, etc. used for data (info) collection in this research.

1.7. Limitation of the Study

The major limitations of the study were a shortage of academic studies about the regulation of AI in Ethiopia in general. The lack of adequate published researches and experience in relation to the concept of AI, by itself it is not as such familiar, because, it is a newly introduced and emerging technology at the national level.

1.8. Scope of the Study

This study was delimited under the thematic range of the “Artificial intelligence and the challenges posed by the nonexistence of ex-ante regulation in Ethiopia. In undertaking the research the relevant government agencies would be taken into consideration, especially those institutions in the federal ministries in applying a qualitative inquiry.

1.9. Organization of the Chapters

In chapter 2, the research will strive to make the foundation for answering the core question to be raised in the research. So, as the question revolves around legislation issues on an ex-ante basis, it tries to briefly examine its definition and nature. It will analyze the existing legal frameworks in Ethiopia. In this regard, it will make a general overlook of what are the advantages and disadvantages of preparing legislation on AI.

In chapter 3, it will take the discussions about regulation held in the previous chapter and make a cross-reference to their applicability in the arena of the AI. By doing so, an attempt would be made for the assessment of risk in preparing legislation for AI in Ethiopia. This will result in an extensive risk assessment, analyzing the benefits and drawbacks a regulation might have on artificial intelligence, and how it reflects on our society. With such analysis, the chapter will begin to answer the main question, the underlying rationale of ex-ante regulation of Artificial intelligence in Ethiopia.

In chapter 4, it still continues to answer the main question of the research. Considering the premise those fundaments this research, this chapter will analyze the steps necessary for intervention on A.I. On the ex-ante basis in Ethiopia to be successful and whether it is possible to achieve such accomplishment. Under the analysis made in this chapter by recapitulating the discussions that are carried so far, so it can confirm the answer to the main question and finally, chapter five will be the recommendation and the conclusion of the research.

Therefore, in sum, this particular research work will do a number of literature reviews for an ex-ante, the legislation of artificial intelligence in Ethiopia, departing from a legislator’s viewpoint. By doing so it shall analyze and attempts to reveal if the benefits of regulating AI outweigh its risks before the fact, with the purpose of protecting the interests of the Ethiopian society. This methodology will be utilized to the achievement of the necessary knowledge to reach a satisfactory conclusion and produce original work to complement the existing literature.

1.10. Ethical Consideration of the Study

Free consent, Confidentiality of the participant will be considered, the participant confidential information would be kept in secret, Anonymity of participants respected, unless the consent is given by the participant, their identity would not be disclosed; protecting intellectual property right, Proper acknowledgment would be made for the contribution of the study.

Chapter Two

2. Conceptual and Theoretical Framework of Artificial Intelligence

2.1. Definition and Nature of Artificial Intelligence

Before defining and dealing with the topic of artificial intelligence, we need to define the word intelligence in the first hand. The term Intelligence in humans has been described as a set of factors that contains as a part: consciousness, self-awareness, language use, the capacity to learn, abstract, adapt, and the capability to reason. Once the intelligence is defined, estimations or approximations of those distinctive attributes or characteristics should form the benchmark of an attempt to create or simulate artificial intelligence. But the question is which characters of the AI can state a simulation to be called AI.35 John McCarthy who is the computer scientist, from the beginning coined the term AI, did not limit intelligence in AI to the replication of human intelligence but argued that machines could show additional intelligence that includes "much more computing than people can do. “He defined AI as "the science and engineering of creating intelligent machines, in particular intelligent computer programs." 36

Additionally the Oxford Dictionary describes artificial intelligence as the “theory and development of computer systems which have the capacity to carry out tasks in a normal manner depending on human intelligence,” A definition that is also used by some scholars, However, others prefer to define it as “the design of intelligent agents,” including different kinds of intelligence, not only human. It is also a development of a flexible agent, which is capable to make suitable it to the new environment not previously familiar and learning through experience, accomplish a goal not possible for traditional computer systems.37

An idiomatic expression of the phrase “artificial intelligence” is an umbrella term for technologies that appeared to act as if they were rational beings Employing some combination of algorithms or a process or set of procedures to be followed in calculations or other problem solving operations, especially by a computer in other term “machine learning,” sensory feedback systems, and automation which means the use or introduction of automatic equipment in a manufacturing or other processor facility. AI technologies in a simple manner carry out the functions that they were programmed to learn. Yet, to the human notice, it appears as if a machine or computer did not perform the activity at all. It looks like as if absolutely or completely a Human is behind the scenes. A human is not, but this clever kind of programmed labor can provide remarkably large advantages or profits to society.38

AI gains positions in new industries and becomes further involved in our day-to-day lives, and that tendency seems likely to continue for the predictable future. The potential for more rapid improvements in AI technology has encouraged expressions of alarm from many adjustments, including some calls for government regulation of AI development and control on AI operation. That is hardly surprising; fear of technological change and calls for the government to regulate new technologies are not new phenomena.39

Artificial intelligence is the general name of the technology for the development of machines, which are brought into existence completely by artificial means and can exhibit performances and behaviors which are like human beings, without taking advantage of any living organism.40 Since it is difficult to describe Intelligence by itself, defining AI would also be problematic. This could be one of the reasons why there are different definitions of AI in the literature found from various sources. There are obviously different opinions on the subject of concern of what should be included in a definition of AI, and what should not be.41

2.2. Philosophy and Theory of Artificial Intelligence

Intelligence is a manifestation of rational action; an intelligent agent takes the most excellent or desirable action in a given situation. Philosophy form a concept of this idea, which later formed the underlying support of AI, by equating the behavior of the mind to that of a machine, it operates on knowledge changed into a coded form in some internal language, and that action or process of thinking can be used to pick out what actions to take.42 The question of Can AI competes with Biological Intelligence? Will obviously have a different answer depending on when the question is asked. It is only possible to perform better than the capacity of the human in some cases, the more variables that are added to a problem, the harder it becomes for humans and at a certain point, and the computer starts to outperform them. AI is different, and therefore not always comparable to, biological intelligence.43 Dr. Martin accepts that recently discovered intelligence is not similar to human intelligence, that in his article “Alien Intelligence” in the “Journal of Business Strategy”. He used a new substitutive term for artificial intelligence which is called "alien intelligence,”.44

2.3. History of Artificial Intelligence

The historical emergence of artificial intelligence officially dates back to 1956 when the pioneer artificial intelligence applications were brought into use for the first time during this period 45. It has been studied for decades and is still one of the most elusive subjects in Computer Science. The term Artificial Intelligence was first coined by John McCarthy in 1956 when he held the first academic conference on the subject. But the journey to understand if machines can genuinely or properly think began much before that.46

The period between 1965 and 1970, could be called a dark period for artificial intelligence. The specified state of growth or development of artificial intelligence in this period is too law to be tested. The hasty attitude due to the unrealistic expectations that become gradually clear and has directed to the idea that it will be easy to uncover the machines with intelligence. But this period was named as a dark period on behalf of artificial intelligence because it did not achieve the aim of creating intelligent machines by merely uploading data.

Even though the time was Dark Age for AI in 1965, Irving John Good put an idea for consideration that the society would be transformed by the creative capacity of a machine with ultra-intelligence. It would be better than human intelligence and intelligent to design even more intelligent machines. He argued, that the latest machine that humans would ever need to create for themselves and would protect humanity. Good argued that this ex machine in the human image would be designed from an understanding of the human intellect. Good's hopefulness and confidence in the future positive success of AI was not shared by subsequent scholars of the time, such as Vernor Vinge, who saw a super intelligent machines not as saviors, but as the arrival of doomsday and after all the machine attained human-level intelligence, it would not remain at that level for long and would reach super intelligence and more extensive quickly. Additionally, Vinge argued that such a machine could have knowledge or perception of its own superior intelligence. This event, which he described as the singularity, would lead the end of humanity. 47

Between 1970 and 1975, artificial intelligence gained momentum. Because the success attained in artificial intelligence systems that have been developed and developed on matters such as disease diagnosis; the foundation of today's artificial intelligence has been settled. From (1975-1980) the indication that they could benefit from artificial intelligence through other branches of science, such as psychology become larger and more advanced.48

Despite the aforementioned limitations in the academic field, the late 1990’s and the beginning of the 21st century saw an occurrence of dawn of hope, which is the engagement of the private sectors to invest in artificial intelligence, a recreational or sporting activity on productivity and in result, the profitability of this new occasion come into existence .49 Thus, some years ago, the world witnessed the action of inserting intelligent mechanisms into people’s everyday lives. From the begging, it began with additions to pre-existing technologies.50 But not time after, these advancements made imaginable the creation of new tools, like virtual personal assistants, home assistants, and recommendation services’51

In the second half of the 19th century, the Industrial Revolution and the development of Capitalism led to a fresh symbolism of science and knowledge which was characterized by technical exhibitions. Therefore, trades and industries created new societies. In this scenario, research methods changed as well and pragmatic research became a priority.52

2.4. The Interplay between Science, Law, and Technology

Nowadays, even though technology grows fast and the capacity to understand technology becomes more complex in order to understand the connection between law and technology, the scientific progress may be not capable to state that what will happen in the future or it is changeable time to time and so it may be out of control but progress allows carrying out suitable or appropriate studies on this issue with the aim to maximize social benefits derived from science. The scientific and technical nature of law shows they are dependent on each other between these poles within the economic and social marked change in nature, form, or the way that law appear.53

Law, science, and technology have a mutual effect on each other; law dealing with the regulation of scientific activity, its products, and scientific knowledge force a way into legal sorts. Particularly, science and technology are used for both safety regulation and regulatory tools taking into account their action of consideration into law.54

It is significant to get the right balance between technology/science and law55 Technology or science has a specific nature that is completely different from law, their development is unpredictable for this reason hard law could slow down the scientific and technological development so that flexibility of law is the necessary factor in any way.56

Regulation has influencing power on the subject's behavior the law is for certain the most example of this without any doubt. Nowadays there is an implementation of an unstated numeric of modern socioeconomic integrative action or influence which needs a change from a command and control legislation and enforcement, a character of the classical democratic setting, to such type of co-regulation and participatory administration of the legislative system that incorporates external players, who are not portion of the State institutions, called to intervene. For this very cause, it looks as if the rigid approach of the ordinary hard law often becomes to be insufficient, whereas soft law may be more desirable.57

The law by itself has to incorporate legislative, executive, and judicial functions on the substantive subject or range of science and technology themselves. The course of a particular reciprocal action or influence between science and law or in another way, whether one is involved in the consequence of law on science and technology, or the effect of science and technology on law and Crosscutting issues. For instance, when we take the intellectual property case, access to research data, proof of identity of expert witnesses, ethical responsibilities of professionals, and the economic special effects of regulation have its own direct involvement in technology/ science. 58

2.5. Principles and Rule

The regulation of AI in one or another way can help consumers to build trust in this new artificial intelligence technology, which is necessary or essential to the healthy growth of the sector. Such trust also has the capacity to increase the use of this technology for the building of a better society. To build such confidence, however, the public at large must ensure that they are protected from and not exposed to danger or risk of AI.59

When we raise the concept of rules and principles in relation to this different technology a number of questions seemingly existent: the first one is the interests of AI regulation protects, secondly, whether the existing regulatory structures are adjusted to new condition; thirdly The regulatory burdens be kept proportionate, finally the exact title role of government to play. In this regard, for instance, the US October 2016 Future of AI report sets out risk dependent public protection and economic fairness as the key regulatory interests, using current regulation as the start point where possible and not stifling innovation: it is obvious that AI has applications in many products, for example, cars and aircraft, which are issue to regulation designed to keep the public safe from harm and to make certain fairness in economic competition. Generally, the way of dealing with regulation of AI-enabled products to protect public safety should be based on a sound understanding of the fact by assessment of the features of risk that the addition of AI may decrease alongside the aspects of risk that it may increase. If a risk falls within the boundary of an existing regulatory regime, moreover, the policy discussion should start by considering whether the existing regulations already sufficiently address the risk, or maybe they need to be adjusted to the new condition of AI. Policymakers should consider how those reactions could be adjusted to lesser costs and barriers to innovation without adversely impacting safety or market fairness.”60

To accomplish the best results, policymaking must, for the most part, more than anything else should focus on getting to know artificial intelligence and make ready the societies for it. Periodic discussions, constant evaluations, public hearings, and steady scientific studies must be certain. Social and economic policies need to be discussed to make sure that society is equipped for the implementation of A.I. Then, any regulation must manage or handle the responsibility and liability issues. Multi-stakeholder reunions and audience discussions must occur so that a complete analysis can be carried and a right balance between consumer protection and industrial profitability can be able to touch. 61

The next step is to make sure that artificial intelligence or this new technology approves to recognize socially relevant values and the inclusion of principles that will serve as a note to any rule and will shape the regulation as a whole. In addition to this, to have in mind when discussing the principles that shall govern artificial intelligence is that the data used to feed the algorithms have great significance for Self-learning tools. Therefore, such databases should also be just, complete, unbiased and high quality, abstain discrimination and support or encourage inclusiveness, as well as to avoid accidents and errors.62


1 Hruy Tsegaye, (2013,), An Appeal to Trans humanism on the Question of Technological Inequality in Africa, available at accessed on March 1, 2019.

2.Kevin Hood, (2018), horizon scanning: Artificial Intelligence and the Legal Profession, page 3, available at profession accessed December 17/2018.

3 Infra note 38, page

4.J.K.C. Kingston, (2016), Artificial Intelligence and Legal Liability, University of Brighton, BN2 4JG, UK, (2016),page 2, Available at, accessed at December 19, 2018,

5.John McCarthy, ,(2007), What is Artificial Intelligence?, Computer Science Department Stanford University, page.2 Available at httpwww-formal.stanford.edujmcwhatisai.pdf accessed on December 20, 2018

6 ibid

7.Chris Boyd and Amy L. Halverson Artificial Intelligence in the Law Firm: Implications for Professional Development Priorities and Practices, ,(2017) , available at accessed November 19, 2018, 8:31:16 PM

8. Supra note 2 page 3

9.Dr Jacob Parakilas, Mary L. ‘Missy’ Cummings , Dr Heather Roff, Kenn Cukier and Hannah Bryce, 2018 ,Artificial Intelligence and International Affairs,available at intelligence-and-international-affairs#pdf accessed on April 24/2019

10.European Consumer Consultative Group opinion, (2018), policy recommendations for a safe and secure use of artificial intelligence automated decision-making, robotics and connected devices in a modern consumer world, availableat partners-consumer-issues/european-consumer-consultative-group-eccg_en.Accessed on December 16/2018.

11. Paper adapted by the Web Foundation from a draft report commissioned to Cath Elliston, Matt Fenech, and Olly Buston of Future Advocacy, (2017), Artificial Intelligence, the Road Ahead in Low and Middle- Income Countries page 4. Available at accessed on 10 March, 2017.

12.Prime minister Dr. Abby Ahmed, Ethiopian Prime minister Dr Ably Ahmed surprise speech, Ethiopian national theater ,the location at Ethiopian national theater,28 January, 2019.transleted by researcher.

13.The Road Ahead in Low and Middle-Income Countries 2017 ARTIFICIALINTELLIGENCE, page .4, available at countries/.accessed on April 24/2019

14.João paulo De Almeida Lenardon (2017), The Regulation of Artificial Intelligence , page 17 available at httparno.uvt.nlshow.cgifid=142832 accessed on December 20/2018.

15 ibid , page 18.

16 ibid, page 30.

17.Maria Assunta Cappelli,(2015), Regulation on Safety and Civil Liability of intelligent autonomous robots the case of smart car A.A. page 11 available at accessed on December 18,2019.

18.supra note 10 page 12


20 ibid

21.Supra note,, 14

22.Richard Kemp, (2016) Legal Aspects of Artificial Intelligence, page .2 available at 2018.pdf accessed on December 18, 2019.

23.supra note 11.tables -1 page 6.

24.ibid page.9

25.ibid page 6

26 ibid page 4.

27.supra note 10. Page 12

28. ibid

29.Kuo L, 16 (2016), ‘Only China, Syria, and Iran rank worse in internet freedom than Ethiopia’, available at and-china/, accessed on 15 March 2017

30.Matthew U. Schere, (2016), Regulating Artificial Intelligence Systems: Risks, Challenges, Competencies, and Strategies, Harvard Journal of Law & Technology Volume 29, Number 2, page 360, available at accessed on December 19, 2018

31.Shruthi Anand, Artificial Intelligence–Literature Review, available at governance/files/artificial-intelligence-literature-review accessed on , December 19, 2018

32.supra note

33.Supra note 17, page 11.

34. Supra note 11, page 4.

35.Michael Guihot, Anne F. Matthew & Nicolas P. Suzor, 2017, ARTICLE: Nudging Robots: Innovative SolutionstoRegulateArtificialIntelligence,page5,availableat ?abstract_id=3017004 accessed on January 26, 2019.

36. ibid page 5

37.supra note 14 , page 7.

38.Adam Thierer, Andrea Castillo O’Sullivan, and Raymond Russel,( 2017), Artificial Intelligence and Public Policy, pp. 5-6,available at:, accessed on January 26, 2019,.

39.Supra note 30, PP 354-355,

40.Yapay Zekânın Tar and Maad M. Mijwel (2015), History of Artificial Intelligence. University of BaghdadBaghdad,Iraqpage2Availableat ects/history-ai.pdfaccessed on, February 7, 2019,

41.Tomas Eric Nordlander,( 2001) ,AI surveying: artificial intelligence in, page 17, available at at , February , 2019,

42.supra note 36, page 1.


44.ibid page 17.

45.supra notes 40 page2.

46.Chris Smith and Brian McGuire, (2006), History of Artificial Intelligence, page 4 available at httpsmafiadoc.comthe-history-of-artificial-intelligence_59818cb51723ddf256291000.html, accessed on, February 8, 2019.

47.Supra note 35, page 9.

48.supra note 40, page2.

59.supra note 14 , page 1.

50.James Wexler, (2002), Artificial Intelligence in Games: A Look at the Smarts behind Lionhead Studio’s “Black and White” and Where It Can and Will Go in the Future’ (dissertation, University of Rochester) available at> accessed 30 November 2018;.

51.supra note 14, page 8, João paulo De Almeida Lenardon (2017),

52.supra note 17, page11, Maria Assunta Cappelli, (2015),

53.Steven Goldberg, (1995), The relationship between law, science and technology availableat,accessed on March 19, 2019.

54.supra note 17, page 11.

55.Erica palmerini, (2013), the interplay between law and technology, or the RoboLaw project in context,LawandTechnology.,page15:availableat lmerini_Intro.pdf,accessed on March 19, 2019

56.supra note 17, page 11.

57. Supra note 56, page .15.

58.A convergence of science and law. a summary report of the first meeting of the science, technology, andlawpanel,nationalacademypresswashington,d.cavailabeleat ter/1#iii.accessed on march 19, 2019

59.Jill Priluck, (2015)‘When Bots Collude’ the New Yorker (New York).available at > accessed 15 April 2018;

60. Supra note, page 22.

61.Supra note 14 page 8.

62. ibid.

Excerpt out of 92 pages


The ex-ante regulation of artificial intelligence in Ethiopia
Ethiopian Civil Service University  (leadership and good governance)
very good
Catalog Number
ISBN (eBook)
Quote paper
Hiwot Degu (Author), 2019, The ex-ante regulation of artificial intelligence in Ethiopia, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: The ex-ante regulation of artificial intelligence in Ethiopia

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free