Concerns
The proliferation of AI technologies in Africa is not without its challenges, as foreign companies and governments have been known to introduce technologies that have questionable benefits for local communities. China has been at the forefront of exporting AI-driven technologies to Africa, with its Belt and Road Initiative and Digital Silk Road Initiative paving the way for introducing smart city infrastructure, 5G networks, surveillance cameras, cloud computing, and e-commerce in many African cities.
The impact of AI technologies developed elsewhere without meaningful testing, impact assessment, or local skills development can be detrimental to fundamental human rights and political stability. A few multinational monopolies, which can undermine local businesses and domestic growth potential, typically develop and offer AI technologies. Huawei and AWS are among the companies that dominate the market, with the latter offering cloud services that are essential for AI systems and the storage of large datasets. However, establishing highly technical cloud services and the massive data centres needed to house and power these data-driven services requires substantial resources that are often only accessible to Big Tech.
In South Africa, AWS is currently constructing a new data storage centre in Cape Town as part of its regional headquarters strategy. Unfortunately, the 150,000m2 development is on contested indigenous land of South Africa’s first peoples, the Khoi and San, who were dispossessed of the area by Dutch colonizers in the 17th century. Despite campaigns against the development and an ongoing court case, the promise of generating 8,000 direct and 13,000 indirect jobs is a compelling counterargument in a country where the official unemployment rate is 46.6%, according to the expanded definition of unemployment following the COVID-19 pandemic.
As the world becomes increasingly digitized and data-driven, artificial intelligence is transforming the global job market. This shift is notable in Africa, where historically, multinational corporations have sought cheaper labour. Today, the rise of AI has created a new form of insecure and precarious employment, which is often extracted from impoverished communities. This type of work is referred to as micro digital labour or “click work’’.
Click work typically involves labelling vast amounts of data and requires workers to have access to a computer. Workers often must bid for such work and are paid a fraction of a penny per unit of work. Many click work assignments involve emotionally distressing content, such as labelling violent and disturbing images or videos. Recently, there have been investigations into the use of refugee camps as a source of cheap digital labour. For example, in Dadaab, one of the world’s largest refugee camps in Kenya, hundreds of computers and yards of wires have been set up to facilitate click work for camp residents.
It is important to recognize that the demand for click work is driven by the global AI economy, which requires vast amounts of labelled data to improve machine learning algorithms. Unfortunately, this demand has led to a situation where vulnerable populations are exploited for cheap labour, without proper protections or benefits. There is a pressing need to address the ethical and social implications of the AI-driven labour market, particularly in regions where poverty and inequality are widespread.
Digital ID systems, which often incorporate biometric technologies like facial recognition, fingerprints, and iris scans, are becoming increasingly popular in Africa. Governments and businesses are adopting these technologies to centralize and streamline government services to prevent fraudulent claims, as well as to protect consumers from identity theft. However, there are concerns regarding the collection and protection of personal data, and the potential for misuse and abuse of this data.
MTC, Namibia’s leading internet services and telecommunications company, is implementing an AI-driven digital ID verification system to improve customer access to its services. This system will collect and store vast amounts of personal data, including facial images and fingerprints, which may be linked to other sensitive data like geolocation. But there is little information available about how this data will be protected or who may have access to it.
In South Africa, an AI-driven digital ID system was used to provide access to social grants for grant recipients. This system was problematic, as the company responsible for distributing the grants, Cash Paymaster Services (CPS), had access to the personal data of all 18 million beneficiaries. CPS used this data to offer predatory financial services to these vulnerable individuals, resulting in many of them receiving little to no grant payment each month. This case highlights the unethical sharing of personal data and the lack of awareness of digital and information rights in South African communities.
Facial recognition technologies, another form of biometric AI, are also being used within digital ID and biometric systems in Africa. Despite this, these technologies are controversial due to inaccuracies, particularly for women, gender minorities, and non-white populations. Facial recognition systems developed elsewhere may also misread African faces and limit human rights, such as freedom of movement, association, and the right to equality and fair treatment. For example, in South Africa, Vumacam has established a network of surveillance cameras using a Danish-built facial recognition system to profile suspicious behaviour. These technologies may be at odds with the country’s democratic vision of an equal society. In Uganda, for instance, Huawei’s AI-powered facial recognition system was used in the 2020 elections to identify, track, and arrest supporters of the opposition wing, which raises concerns about privacy and fundamental human rights violations.
The above analysis suggests several key policy areas for African policymakers and technologists to focus on. To kick start, it is essential to prioritize the development of safe and inclusive infrastructure to support the local development of technology. This includes policies to improve internet access and prevent internet shutdowns, as well as policies to support good governance and availability of data for development. African Union Data Policy Framework and Africa-EU Global Gateway should promote uniformity in data governance standards across the continent and establish equal access to basic and advanced digital infrastructure.
Besides that, policymakers should consider the benefits of regional cooperation for developing common regulatory responses to multinational and foreign tech companies operating in the region. Cooperation between countries could also extend to developing taxation provisions for multinational social media platforms and data-sharing agreements to provide access to a broader range of public sector data. The African Continental Free Trade Area Protocol on E-Commerce may present an important opportunity for integrating a provision to support inter-regional data sharing, supporting regional development goals, and economic growth.
To conclude, policymakers should prioritize the development of local capacity and skills in technology-related fields, such as AI and data science. We need capacity development policies to promote understanding of AI at all levels, with specific measures to advance women in STEM and AI-related decision-making positions. Policies may also include measures to attract diverse AI talent by lifting entry barriers between countries for Africans with data science and computing skills. By focusing on these key policy areas, African governments can ensure that AI adoption is inclusive and does not perpetuate social inequality while promoting sustainable development and growth across the continent.
Author: Osei Manu Kagyah
The writer is a member of the Institute of ICT Professionals Ghana (IIPGH) and works at the nexus of technology and society as a Technology Policy Advocate / Analyst. For comments, contact kagyahosei@gmail.com