Ethics are moral principles that govern a person’s behaviour or the conduct of an activity. As a practical example, one ethical principle is to deal with all of us with respect. Philosophers have debated that ethics for plenty centuries and there are numerous famous concepts, perhaps one of the most well-known being Kant’s express imperative act as you may like every other humans to behave towards all different humans (From Kant’s 1785 ebook ‘Ground work of the metaphysics of morals’).

Different Artificial Intelligence (AI) ethics standards, guidelines and strategies have been formulated round the world to assist remedy the rising issues of Artificial Intelligence structures in our society, like OECD (Organisation for Economic Co-operation and Development), AI4People’ Ethical Frameworks, The Beijing AI Principles. Application of responsible AI in finance, area exploration, superior production, transportation, electricity improvement, Agriculture and health is a crucial aspect of growing AI systems.

We have noticed the effect of automation on “blue and white-collar” jobs, but as computer systems turn out to be very complex and useful, more jobs and positions are made obsolete. The effect of AI and robotics in Africa will lead to inequality; there will be a large quantity of unemployable humans and breakdowns in the social order (Smith and Anderson 2014). The most important question around AI is inequality, which isn’t normally covered within the debate in AI ethics. It's an ethical problem, however, it’s mostly an issue of politics ‘who benefits from AI’ (Stilgoe 2020). It has been observed that lack of government engagement to date has been a hindrance and encouraging African governments to take a proactive approach to AI policy.

In reality, possible dangers emerge from the AI race narrative, in addition to a really competitive race to develop AI systems for technological superiority (Cave and ÓhÉigeartaigh 2018). In drafting these legal guidelines, African regulators should learn from International best practices, which include warding off burdensome requirements which could foreclose the benefits of AI and position African businesses at a disadvantage. Ethical techniques should remove bureaucracy that will mitigate and intervene in the long standing societal biases, mainly on the grounds of protected characteristics such as race and culture. Developing African-Centric AI models, the use of information sourced from Africa will make it easy for practical implementation of theoretical AI ethics in Africa, although UNESCO survey results shows that African countries like Congo, Sao Tome and Principe, Zimbabwe have developed ethical guidelines for AI but a general and harmonized approach is necessary to protect individuals and collective privacy rights in the cross border of AI data.

1 Ethics in Artificial Intelligence

There ought to be ideas that govern the discovery and design of Artificial intelligence agents. AI is defined as developing machines that could assume, learn and react like human (Siau and Wang 2020). We will similarly broaden the definition to describe machines that do things that could normally require human intelligence—things consisting of speech recognition, visible perception, and choice making (Siau and Wang 2020). The conduct of the scientists inventing these machines and the machines themselves can also have moral issues. In a recent managed study, researchers reviewed those moral recommendations function as a foundation for ethical decision-making for software program engineers (McNamara et al. 2018). There is poor practice of AI systems in compliance with the implementation of ethical guidelines, although there are some AI ethics frameworks formulated from different parts of the world today, namely.

1.1 OECD (Organisation for Economic Co-operation and Development)

This AI ethics framework recommended five complementary values-based principles for the responsible stewardship of trustworthy AI, which includes that AI should drive inclusive growth, sustainable development and wellbeing; AI systems development should consider rule of law, human rights, democratic values and diversity, appropriate safeguards; AI system outcomes should be transparent and properly disclosed to the users; Safety and security should be incorporated during the development plan of AI systems, finally organisations and individuals developing, deploying AI systems should be accountable to the outcome of the AI systems.

1.2 AI4People’ Ethical Frameworks

This Artificial Intelligence ethics guidelines is about Human organisation and oversight which include monitoring, schooling, human-machine interfaces and external control of automobile data; Technical robustness and safety—together with resilience to assault and security, fall back plan and regular safety, accuracy and reliability; privacy and data governance together with respect for privacy, transparency & communication, and access to facts; Transparency as a key mechanism to comprehend all other requirements; Diversity, non-discrimination and equity which includes the avoidance of unfair bias, accountable balancing and accessibility; Societal and environmental wellness—including sustainability and environmental friendliness and social effect; Accountability—which includes audit ability, measures of transparency, reporting of poor effect.

1.3 The Beijing AI Principles

The creation of Artificial Intelligence (AI) worries the future of the entire society, all humankind, and the environment. The ideas are proposed right here as an initiative for the research, development, use, and governance and long-term planning of AI, calling for its healthy improvement to assist the development of a network of common destiny, and the belief of useful AI for humankind and nature. The improvement of AI have to do nicely to all mankind and conform to human values like privacy, dignity, freedom, autonomy, and rights, the developers or researchers of AI systems ought to be responsible for the outcome of their merchandise and additionally endeavour to control the activities of the AI systems to divert risk. There ought to be a reflection of diversity and inclusiveness when constructing an open AI machine. Additionally, customers of AI systems have to be taken into consideration so that they need to gain knowledge of how to use the AI systems wisely and have enough information on how the AI system affects their rights and interests. Governance of AI ought to look at Optimizing employment; concordance and cooperation need to be actively increase so one can express the philosophy of “optimizing philosophy”, and long-term planning to encourage Artificial general Intelligence.

Having looked at globally present AI ethics frameworks like OECD (Organization for Economic Co-operation and Development), AI4People’ ethical Frameworks, The Beijing AI principles, it's far clearly beneficial for African countries to form a unifying AI ethics framework for AI implementation to facilitate the practice of AI ethics in Africa.

2 Suggestions for Practicing AI Ethics in Africa

Practice of AI Ethics that is created with African context in Africa will definitely cause independent and complete representation of human beings from different kinds of ethnic agencies, as AI permeates our lives in Africa. Africa is a continent made up of nations with unique cultures and values. There is need to bear in mind the peculiarity of Africa when designing or developing AI Ethics for practice in Africa. The increasing adoption of AI faces certain challenges and constraints, AI and the associated strategies of machine learning, deep Learning, Data science, and so forth is based on getting access to massive quantities of facts that may help train and develop new systems. The variety and value of facts will decide what model will analyze, if the records is western way culture inclined, it will affect the outcome of the system. This imbalance carries dangers, in particular wherein the moral norms and values designed into those technologies collide with those of the African groups in which they're deployed. AI ethics which are applicable inside the context of African Continents are unique to African nations. This can be achieved through engaging in the following activities.

2.1 Defining African Values and Align AI with Such Values

Any AI project implementation should be preceded by values and risk assessment. In doing so, they should choose and contextualize only the good parts of global initiatives that accord with African own ethical values and cultural contexts. What ethics means to Africa may be different to other regions, for instance, African cultures, despite the diversity, have certain commonalities that include Ubuntu which encompasses a collective approach to life, sentimental and religious values and beliefs that include the desire for collective good outcomes while being conscious of evil, as well as a belief in life after death values, may not accord with or may even collide with African interests in the context of AI. African beliefs and values consideration, determine the type of AI that is created.

2.2 Protecting Data Privacy and Privacy Rights in Cross Border Data Flows

AI affects the privacy of individuals especially through big data: long term records that can be kept on any one who produces storable data. Digital records can be searched using algorithms for pattern recognition meaning that we have lost the default assumption of anonymity by obscurity (Selinger and Hartzog 2017). Any one of us can be identified by facial recognition software or data mining of our shopping or social media habits (Pasquale 2015). AI has been making such massive progress for several years precisely because of the large amounts of (personal) data available. Those data are collected by privacy-invasive social media platforms, Smartphone apps, as well as Internet of Things devices with their countless sensors. In the near future, vision-based drones, robots and wearable cameras may expand this surveillance to rural locations and one’s home, places of worship, and places where privacy is considered sacrosanct, such as bathrooms and changing rooms. As the applications of robots and wearable cameras expand into our homes and begin to capture and record all aspects of daily living. We begin to approach a world in which all, even bystanders are being constantly observed by various cameras wherever they go (Wagner 2018). Presently, Africa is adopting AI at the stage of implementation when data is more important than the technology itself. African countries have already been working on data issues and facing data protection challenges in the process while some are in the process of coming up with a legal framework. We can conclude that the current AI boom coincides with the emergence of a post-privacy society. Collective rights of peoples and communities must be protected in addition to personal privacy. This can be achieved through the Harmonisation of Data and AI frameworks in Africa.

2.3 Ensuring Quality of Data and Removing Biases

Building AI systems, most often requires people to manage and clean up data to instruct the training algorithms with huge training data sets, the question is the socio-demographic source of this data determines how the AI model will behave. The inventor or developer of an AI has great potential to determine its use and reach (Conn 2018), suggesting a need for inventors to consider the wider impacts of their creation. AI machines use algorithms (a set of sequential rules to be followed in problem-solving) created by humans, so if the creator has any inherent biases or is judgmental in some way, those biases can be built into the machine. Imagine a machine used to predict criminal behaviour that includes the creator's bias against a particular race. That wouldn't be fair or neutral, Systematic bias may arise as a result of the data used to train systems or as a result of values held by system developers and users. Researchers have found that automated advertisement tools are more likely to distribute adverts for well-paid jobs to men than women (Datta et al. 2015). AI that is biased against particular groups within society can have far-reaching effects. Its use in law enforcement or national security, for example, could result in some demographics being unfairly imprisoned or detained. If AI is used to screen people for job applications or university admissions it could result in entire sections of society being disadvantaged. Using AI to perform credit checks could result in some individuals being unfairly refused loans, making it difficult for them to escape a cycle of poverty. However addressing the gender biases, culture biases and race biases in the development and use of AI systems will ensure Al ethics with African context, will be implemented in Africa.

2.4 Introducing Safeguards to Balance AI Opportunities and Risks

AI offers opportunities mostly in the following areas: research and innovation; smart automation in core areas such as health e.g. disease diagnostics such as malaria and TB test automation and smart involvement and participation, clerical roles, agriculture which employs 70% of Africa’s labor, energy, and tourism. AI also presents opportunities for efficient public sector decisions and resource allocation especially in social protection schemes, business analytics through increased data intelligence. One way of programming AI systems is reinforcement learning where improved performance is reinforced with a virtual reward. If we consider a system to be suffering when its reward functions give it negative input, therefore once we consider machines as entities that can perceive, feel and act, it is no huge leap to ponder their legal status. In practicing AI Ethics in Africa, AI machines should not be seen as moral agency or seen as persons.

2.5 Fair and Socially Responsible AI

AI should be both fair and inclusive; taking the continent’s nuances into account. This can be achieved through Use of open data sets for benchmarking fairness. The use of open data sets with safeguards offers benefits for benchmarking fairness among other purposes. Ethically built and used AI could help promote equality and fairness, but poor or malicious design risks exacerbating existing social problems in new ways. Bryson (2019) argues that giving robots moral agency could in itself be construed as an immoral action, as it would be unethical to artefacts in a situation of competition with us. Fair and socially responsible AI needs to include and embody African values.

2.6 Engaging Local Communities

The ethics of community creation goes over and above the legal requirements to include awareness-raising as a public dimension of policy: understanding the peculiar needs of the people, working that is based on dialogue with communities making sure they are included in the research, public engagement during which they can inform the experts of their unique needs, and not exploiting them or simply ticking the boxes as required by the law.

2.7 Social Inclusion of Africa in the Fourth Industrial Revolution

While AI has the potential to help solve many of humanity’s most pressing problems, for example by creating a world that is less sick, less hungry, more productive, better educated, and better prepared to thwart the effects of climate change, evidence suggests that this promise comes with an escalating global risk of entrenched and amplified social inequality. The uncertainty of AI and some of its negative impact is most likely to be felt in the Global South, especially Africa who may be excluded on both social, economic, and various grounds.

3 African-Context—AI Ethics Framework

African Artificial Intelligence should follow these guidelines.

  • Governments in Africa should include AI ethics practice in the government policies.

  • Awareness index of AI ethics in Africa should be upgraded to a high level. To ensure AI can benefit all and create the most societal value, stakeholders need to have open conversations about the ethical dimensions of this technology and take appropriate actions.

  • African values and morals should be considered when building AI systems because Africans have regard for privacy, especially sex life. So, it is against African values and morals to use sexbots which is one of the products of AI.

  • Africans live communally, so local communities should be engaged while building AI systems that will be used by Africans.

  • Africa AI products should not be available only to a particular set of people due to digital divide, lack of social amenities and communication facilities, this can be resolved by ensuring more investment in technology.

4 Conclusion

In Africa, AI can help with some of the region’s most pervasive problems: from reducing poverty and improving education, to delivering healthcare and eradicating diseases, addressing sustainability challenges, the growing demand for food from fast-growing population to advancing inclusion in societies. AI offers vast opportunities in the area of health, transportation; education, agriculture, etc. The ethical implementation of AI faces some challenges like digital terminologies used differently in different continents, countries, and communities which could be expressed due to diversity in culture and way of life.. The digital divide, which is the gap between individuals, households, businesses and geographic areas at different socio-economic levels with regard to both their opportunities to access information and communication technologies (ICTs) and to their use of the Internet for a wide variety of activities, is also a major issue in Africa, which hinders the use of digital services like AI. As a first step in dealing with these challenges, African stakeholders should work together to ensure AI applications are developed by teams with diverse demographic, gender, ethnic, and socio-economic backgrounds. This is important to avoid cognitive bias. AI innovation will be key to solving many of these challenges as long as it integrates consideration of ethical implications of systems that they build. It’s also true that the risks of AI cause human misunderstanding and overreliance on AI systems, as they also have limitations in the systems themselves. Therefore, developers must always strive to fully and effectively communicate with users and regulators to ensure adequate understanding of the technology, its use, and the risks associated with it.