Leveraging AI in Africa to accelerate progress for children

Regional experts discuss obstacles and opportunities facing the AI industry in Africa

Melanie Penagos
11 March 2020
This work is part of UNICEF's AI for Children project
2 minute read : The reflections in this post are from a regional consultation that took place in Cape Town, South Africa, in February 2020. It is the third in a series of regional workshops designed to develop policy guidance for AI systems that help realize child rights. 

In Africa, what challenges need to be overcome so that AI systems empower and protect children? What changes are needed to make this happen?

These are the questions that policymakers, practitioners and academics from 9 African countries tackled during our February workshop on AI and children in Cape Town, South Africa. Workshop participants confirmed what we already know — the AI research community in Africa has grown significantly in recent years, leading to greater learning and employment opportunities for young people. The topic is also on the minds of many African policymakers and politicians as they consider how AI systems might impede or accelerate their local development goals.

The participants echoed the momentum in the industry as they discussed ways to develop and apply African-made AI solutions to African challenges. A commonly expressed sentiment was that industry leaders should not be thinking of how to create the next Uber or Amazon of Africa, but rather how to foster an enabling environment in which local experts can innovate and problem solve through AI, thereby growing African tech giants from the ground up.

This aspiration is not without its challenges, however. We observed that gaps continue to persist between expressions of interest in AI in Africa and concrete steps towards practical applications where AI and children’s rights intersect. 

AI Workshop Cape Town 2020
UNICEF Office of Global Insight and Policy

Main challenges and recommendations

Gaps in ICT infrastructure: According to the ITU, the African region has the lowest internet usage rates, with 75 per cent of the continent’s population still offline. Of those online, most connect to the internet through mobile subscriptions, which are more affordable and available than fixed-broadband access. In Africa, women are on average 34 per cent less likely than men to have a smartphone and access to mobile internet. Additionally, Africa "remains the only continent whose digital gender gap has widened since 2013." This barrier contributes to low digital literacy and confidence among women. Given these inequities, global AI guidance should aim to improve digital and AI literacy of children and those around them, including educators in parents. Such strategies could include funding mechanisms to enhance AI literacy programs and integrate them early on, prior to higher education.

Low trust in public institutions: Discussions during the workshop revealed that mistrust in institutions persists across the continent. Although there is broad recognition that lack of trust is a worldwide phenomenon, not unique to any one continent or country, there is no universal solution. In practice, mistrust can hamper AI development because people may not want to share data with public institutions, nor trust the decisions made by their AI systems. One of the biggest questions raised by participants was how to generate greater trust in governments. Through these conversations, we discussed how thoughtful, rights-based AI policies can hold governments and industry accountable for user data, including children’s data, and serve to help rebuild public trust.

Siloed policy making: Digital technologies are cross-cutting and therefore demand a range of supporting policies and standards, including those addressing data protection, privacy, employment and digital services. Given that most ministries are working on sectoral issues, intersectoral coordination within governments is necessary in order to develop and implement these policies. Without this planning, there is the potential to create misaligned or inadequate policy and legislative frameworks. However, by sharing evidence, lessons learned, and relevant use cases demonstrating the value of AI for development, policymakers can increase their understanding of these issues and create more informed policies that align AI initiatives with their national development plans.

>> Read: Cape Town Q&A with workshop participants

View or download the full workshop report