🔬 Research summary by Arthur Gwagwa, a doctoral researcher at the Utrecht University Ethics Institute. His work is part of the research program Ethics of Socially Disruptive Technologies, which is funded through the Gravitation program of the Dutch Ministry of Education, Culture, and Science and the Netherlands Organization for Scientific Research (NWO grant number 024.004.031).
[Original paper by Arthur Gwagwa, Emre Kazim, Airlie Hilliard]
Overview: Having previously been excluded from the benefits of previous industrial revolutions, in this paper we explore how the global AI ethics community can include views from Sub-Saharan Africa as a way to improve the terms on which African populations and subpopulations and their concerns are included in the global AI ethics discourses to avoid being excluded from the current fourth industrial revolution. Specifically, we argue that the value of Ubuntu could be of immense value in AI applied normative ethics, particularly toward an inclusive approach for the implementation of the universal AI ethics principles and guidelines.
While there is a consensus about the enormous potential for artificial intelligence (AI) to advance development and solve some of the most pressing challenges faced by Sub-Saharan Africa (SSA), discussions of the ethical challenges that AI will bring to Africa have only just begun. Indeed, much of the discourse concerning AI ethics is centred around the views of the Global North, with there being limited opportunity for regions in the Global South, such as SSA, to participate in these discussions and shape the future of AI. SSA will be disadvantaged on two key grounds: Firstly, since AI can amplify or reinforce long-standing societal biases, particularly those related to characteristics protected under international human rights law, such as race and culture, it is important that other viewpoints are also considered in these discussions. Secondly, as Africans can lack the capacity to access and apply their data, they are less able to develop and implement AI and so miss out on the economic benefits it can bring. Currently there is a lack of commitment from the Global North to addressing historical social and economic injustices.
Exclusion of at the Continental Level
While SSA is made up of a diverse range of countries, they share broad similarities, like their history; aspirations, which are mostly shaped by a common history of European domination , broadly similar post-independent political junctures and trajectories; and a broadly similar communitarian cultural value system ensuring an appropriate ethical and legal framework to strengthen African values. With the onset of the Fourth Industrial Revolution, mostly underpinned by AI, Africans may be excluded from the benefits of AI on the grounds of natural characteristics or protected attributes, including color, language, culture, or race, as a result of the limited or unrepresentative African datasets available for the proper training and application of algorithms or AI applications, like facial recognition software. Indeed, commercial gender recognition tools are much less accurate for individuals with darker skin, being the most inaccurate for dark skinned females and the most accurate for light skinned males. It is, therefore, evident that there needs to be an effort toward greater inclusion in this domain, particularly since the Global North lacks the insight needed to create solidarity in these advancements. First, this is due to the disconnect between the algorithm designers and the communities where the research is conducted or algorithms are implemented. Secondly, governance, including in the AI domain, in liberal democracies of the Global North is mainly focused on protecting autonomy within the individual private sphere unlike in Africa where autonomy is relational.
Exclusion at the National Level
Global corporations, including those working on technology and data, are involved in data-mining activities in Africa that are not just amplifying existing societal tensions but also excluding African subpopulations who represent low-value data. This exclusion is also seen in the uneven access to data, AI, and related technologies, as well as the impact of these tools, which is greatest in marginalized populations. This impact is particularly felt in the least developed countries whose populations sit at the intersection of multiple disadvantages. Populations that lack representation in datasets are often forgotten and the gap between developed and undeveloped countries widens. Therefore, initiatives are needed to increase the fairness and representativeness of data and algorithms and an examination of the values that they embody to facilitate greater inclusion.
Do Current Initiatives Embody Africanness?
The current exclusion of Africa, including its ethical approaches to AI governance, whether intentional or unintentional, means the inclusion debate is still framed from the perspective of the Global North, who developed the technology in accordance with Western perspectives, values, and interests with little regulation or critical scrutiny. As African and South-American countries are not represented independently from the international or supranational organizations that are producing these guidelines, this may present a barrier to implementation of such guidelines but also the deployment of the AI technologies in specific sectors, such as agriculture, where, for example, excessive automation may disrupt the African way of life that revolve around certain customs.
Reshaping the Western Concept of Inclusion and Emerging African Views
The future of the inclusion debate will depend on the ability to develop a global inclusion initiative that draws on the intellectual capacities of both the Global North and the Global South. Specifically, Africans should define what inclusion means to them and how it can be achieved. Efforts toward these dialogues have already proven useful, particularly the workshops of the UN Global Pulse, which were held in Ghana and Tunisia. From these workshops emerged a unanimous consensus that Africa could learn from the Global North’s mistakes to ensure that they do not develop technologies without first formulating a set of values to guide them. In addition, Africans advocated for the need for human control of technology and the promotion of human values, something which has been reactionary rather than proactive in global principles.
The Significance of Ubuntu as a Universal African Value
In Africa, or at least in southern Africa, the Zulu term Ubuntu (“a person is a person through other persons”) has been used to describe African morality and way of life. A collective rights approach to AI ethics could be beneficial to the rest of the world if Africa’s Ubuntu ethics and the normative principles emerging from it are incorporated into the global AI ethics discourse. Doing so would mean that technology would be more reflective of the value of life or communion and communal relationships, which are characterized by identification with others and exhibition of solidarity with them.
Between the lines
Ubuntu is not just a basis for communities but can also be a basis for the inclusion discourse and be something Africa can export to global forums in proposing how the benefits of AI can be shared. There is a significant role that African philosophers can play in developing new theories and methods that are necessary to understand, morally assess, and intervene in the development and implementation of AI. This would promote the inclusion of values like harmony, consensus, collective action, and common good, which are characteristic of African normative ethics, in the global discourse and AI policy. Such values would promote inclusive discussions, with the communities likely to be hit hardest by emerging AI technologies given an opportunity to voice their concerns, resulting in AI being more accessible and having less adverse effects for marginalized populations.