The rapidly developing set of Artificial Intelligence (AI) technologies has the potential to solve some of the most pressing challenges that impact Sub-Saharan Africa and drive growth and development in core sectors:
• Agriculture will be done more efficiently and effectively, raising yields.
• Healthcare will be better tailored, higher quality, and more accessible, improving outcomes.
• Public services will be more efficient and more responsive to citizens, enhancing impact.
• Financial services will be more secure and reach more citizens who need them, expanding access.
Forward thinking policy-makers, innovative startups, global technology partners, civil society groups, and international global stakeholders are already mobilizing to promote the growth of a vibrant AI ecosystem in Africa.
However, there remain structural challenges that can hamper the development of a healthy AI ecosystem in Africa:
• Education systems will need to adapt quickly, and new frameworks need to be created for workers and citizens to develop the skills they need
• Broadband coverage will need to expand rapidly — specifically in rural areas — in order for all citizens and businesses to reap the benefits.
• Ethical implications regarding the fair, secure, and inclusive use of AI applications also must be addressed through collaboration and engagement to ensure AI systems earn trust.
• Ensuring a deeper, broader, and more accessible pool of data is available will also be key to enable researchers, developers, and users to drive AI.
• As with other transformative and revolutionary technologies, there are challenges inherent in the development of AI.
• Governments can embrace these challenges and benefit from AI by creating clear roadmaps to guide the adoption of this technology.
• They should recalibrate their laws and legal frameworks to support data-driven technologies and innovation-driven growth; strengthen the supporting infrastructure for development; and set the tone of a collaborative approach that allows all stakeholders to share their expertise, insights, and build trust.
• With the correct mix of policies, Africa and its citizens can reap the benefits of the transformations in the years to come.
The following information is excerpted from the Artificial Intelligence for Africa whitepaper, with permission from the authors at Access Partnership.
The entire, informative paper that discusses the opportunities and challenges for AI in Africa may be downloaded at this direct link:
Government (Public Services) citizens’ experience with public services can often be challenging. Delivery is characterized by backlogs; redundant tasks; lack of accuracy; slow response times; and generally poor quality, which leads to low levels of citizen satisfaction.
Governments’ ability to ensure efficient use of resources in the delivery of public services is impaired due to factors such as corruption and lack of transparency, as well as public service delivery modes that struggle to respond to present day needs. Thus, while delivery of public services commonly accounts for a large proportion of government budgets, increased spending is often not matched by improvements
Through automation, AI technologies can significantly streamline process es and reduce costs: it can ease administrative burdens, paperwork, and backlogs, increasing public sector efficiency and the speed at which public services can be delivered. This will allow public sector managers to resolve resource allocation constraints, redirecting the staff where they can be most productive.1
In addition to making the delivery of existing services more efficient, AI will drive innovation, enabling new and better types of public services. AI’s predictive capabilities are a game - changer for how government services and policies can respond to society’s needs: from pre-emptive social service interventions to help children and struggling students to better crime reporting and emergency response.
AI tools can also better administer infrastructure, anticipating the need for repairs and better managing cyberattacks that threaten critical systems.2
Finally, AI tools will enhance citizen participation. They can provide new platforms for citizens to assess the quality, adequacy, and effectiveness of public services as well as express their needs and preferences. This provides government with more information to improve their services and make more informed policy decisions.3
African governments could help demonstrate AI’s potential benefits and foster public trust in the technology.4 Governments can also help the development of the local AI industry by seeking solutions that use AI to address various governmental needs and thereby creating demand of this technology.
AI Solutions for Government
The Microsoft CityNext Partner designed SANSTAR for the Los Angeles Bureau of Sanitation (LASAN) using Microsoft Azure cloud services.
The smartphone application is used by truck drivers to map and record their daily routes and by citizens to report clean-up issues. The mobile app allows drivers to complete their routes faster and respond to more customer requests.5
This platform enables federal, state , and local government agencies in the United States to seamlessly and automatically process, transform, and analyze their data. The discovery of mission-critical information is expedited across content silos by cognitively processing and searching for faces, objects, spoken words, logos and more.6
Many governments in Africa have begun to take steps to promote AI in their countries:
The government of Nigeria has taken steps to promote partnerships and stakeholder engagement towards leveraging AI’s benefits. The Ministry of Science and Technology has announced the formation of a National Agency for Research in Robotics and Artificial Intelligence (NARRAI).7
The new institute will collaborate with international research bodies, enhance instruction on AI topics for thousands of students, and promote Nigeria’s ability to leverage these technologies for economic growth. In March 2018, Minister of Communications Adebayo Shittu also restated his ministry’s commitment to support AI stakeholders, engage in conversations to manage and explore the implications of AI, and share best practices.8
Kenya was the first African country to launch an open-data portal to make information on education, energy, health, population, poverty, and water and sanitation, which was previously very hard to access, available to citizens.9
Application development in Kenya is high, and the government wanted to support the industry’s growth. The open-data portal was created in response to requests for data by local tech incubators and co-working facilities for Nairobi programmers, such as iHub10, which led the government to recognize that access to public datasets is crucial for developing locally relevant AI solutions and services. So far, data from this governmental portal has been key in the development of about 100 apps.
The South African Department of Trade and Industry formed a Chief Directorate for Future Industrial Production and Technologies (FIP&T) in 2017 to examine the impacts of emerging digital technologies, including the Internet of Things, big data, AI, robotics, and new materials.
The unit aims to build government capacity to address these challenges and partner with industry to enhance South Africa’s readiness.11 Science and Technology Minister Mmamoloko Kubayi-Ngubane has also said that the government aims to boost its investment in research and development, support for entrepreneurs, and skills development12.
Reaching All Citizens
AI depends on high quality broadband. This creates an obvious problem for Africa: given the continent’s many connectivity challenges, people must be brought online before they can fully leverage the benefits of AI.
There are an estimated 267 million individuals not using the Internet in Africa, and approximately 53 million households13. Within these numbers, there are substantial inequalities: while approximately 22 percent of urban populations have access to the Internet, this number falls to just 10 percent for rural populations. There are also similar divides between men and women, the youth and mature populations, and upper versus lower income groups14. Without sufficient connectivity, entire regions will be excluded from all that this technology can to offer.
The inter-related issues of insufficient infrastructure and lack of affordability are the key obstacles. An estimated 30 percent of Sub-Saharan Africa’s population is beyond the reach of even the backbone network, let alone last mile links to access the Internet15.
African countries also have the most expensive broadband in the world. According to the Alliance for Affordable Internet (A4AI), the first gigabyte of mobile data cost 9.3 percent of the average income in 2016, compared to 3.7 percent for Latin America and 2.5 percent for Asia. Of the 59 developing countries studied, African countries occupied 9 of the 10 lowest slots regarding affordability, with costs ranging from 12 percent to 44 percent of income16.
Governments should seek to develop and implement policies to enhance connectivity and affordability. This is especially urgent in rural areas, where the lack of broadband is most pressing, but also where applications in the agricultural sector have significant potential.
While in some cases public support and direct investment may be needed, private investment is critical to ensure adequate infrastructure is in place and should be encouraged.
Enabling new wireless technologies may also be an important piece of the puzzle to connect users as quickly and cheaply as possible, bypassing in some cases expensive terrestrial infrastructure.
Government’s should improve the accuracy of data they collect and share in key sectors — for example, with the use of technology (such as cloud computing) by national statistical agencies to improve efficiency in the gathering, structuring and analysis of data. However, policies must also encourage the private sector and civil society to do the same.
Governments can help support this by encouraging voluntary adherence to industry standards for data that facilitate interoperability of data sets, promoting data publishing principles (for example Sir Tim Berners-Lee’s ‘five-star maturity model’17 and incentivizing automated collection and sharing of certain non-proprietary data by the private sector.18
Encouraging data sharing platforms, for example for publicly funded academic and scientific research institutions, will also help deepen the pool of data available to all stakeholders.
Regulatory and Policy Framework
Policy can be a powerful tool for African governments to promote technological development by encouraging innovation and investment. At the same time, as leading countries have shown, government engagement and experimentation with nascent technology can also be a powerful signal of trust and support local companies. According to African stakeholders, low government engagement, particularly at the policy level, has been a hindrance, and a stronger focus will encourage an early adoption of AI.19
African governments should take a proactive approach and implement AI-friendly regulation, policies, and initiatives. There are several areas relevant to the development and AI and robust digital economies where policy - makers should focus:
• Data privacy and security — A data privacy and security framework that individuals can trust encourages and empowers them to use AI-based solutions that require their data to work. Data privacy and security laws should aim to protect users’ data without restricting the ability to move data across borders. In drafting these laws, African regulators should look to learn from international best practice, which includes avoiding burdensome requirements which would foreclose the benefits of AI and put African companies at a disadvantage.
• Cybersecurity — African governments should adopt cybersecurity laws that provide for meaningful deterrence, incentivize investment, clarify legal responsibilities, and create effective and reasonable enforcement mechanisms. Additionally, authorities should help users understand and properly manage the risks inherent in using AI technology.
• Digital strategies and cloud adoption initiatives — Governments should develop national digital strategies and policies that foster widespread cloud adoption to democratize the use of advanced technologies.
• Intellectual property — Intellectual property laws that provide for clear protection and enforcement against misappropriation and infringement of technological developments, including proprietary algorithms, are indispensable to promote continued innovation and advancement in AI.
• Procurement policies — Public procurement regulations should enable the use of AI solutions for the provision of public services. By investing in public sector innovation, African governments will demonstrate their trust in
AI and support the growth of local developers.
• Industry-led standards and international harmonization of rules — IT organizations worldwide are developing international standards to ensure data portability, interoperability, and a smooth data flow. African governments should remain abreast of developments in this area and seek t o adopt international standards and harmonization rules as they become applicable.
By drawing from international best practice, African governments can develop national strategies and create a flourishing legal environment for AI. Leading actors — both other governments and the private sector — have valuable experiences to share regarding the advantages, uses and risks of AI, as well as policy approaches they have taken to address challenges and create an environment to fully benefit from
Creating a Collaborative Environment
Following the example of leading countries in AI, African governments should increase cooperation and exchange of information between diverse stakeholders: academia; industry (including startups and entrepreneurs), civil society (including NGOs and think tanks); policymakers and regulators. These actors must work together — not in silos.
Such a collaborative approach encourages the sharing of expertise and perspective on AI. African government can gain a deeper understanding on the technology and rely on international best practice to address specific local and regional needs. This approach ensures that policy and regulatory action protects citizens and supports the technology’s development.
Governments worldwide are concerned about the major intellectual, technological, political, ethical, and social questions that will arise as AI become deeply integrated into our lives.20 Local, national, regional, and international collaboration can address these concerns as governments share knowledge and experience.
Collaboration is as much an orientation towards openness as it is a specific policy program. African governments can pursue many types of measures to foster the local and regional AI industry, including:
• Integrate national and regional AI Councils with leading figures from industry, academia and government to advise on the development of AI strategies and oversee their implementation.
• Adopt open data initiatives as a way of using technology to support distributed innovation, and to make AI development more participatory and transparent.21
• Develop frameworks that enable government and industry to work together in technology projects, including public-private partnerships.
• Create national and regional “AI labs” that gather top researchers and thought leaders working in the development of the technology as well as on its implications for policy and law- making purposes.22
• Adopt policies in partnership with universities and the private sector to attract and retain people with AI skills.
• For example, African governments could:
≈ Work with industry and academia to develop and run specialized fellowships and programs (such as Masters in AI) that prepare the national workforce to respond to businesses current and future skills needs.
≈ Create incentives for industry to fund educational programs at all levels, from basic skills to masters and doctoral programmes.
≈ Consider modifications to visa allowances to allow foreign nationals to work in jobs in science and digital technology.23
≈ Work with other stakeholders to organise networking events that allow experts to share knowledge and collaboration across countries. Some countries in the region are already doing this, such as Nigeria (Data Science Lagos), Kenya (Data Science Nairobi) and South Africa (Machine Intelligence Institute of Africa).22 These will also be key to ensure Africa’s young, ambitious, and entrepreneurial-minded population find what they need in Africa, rather than go to other countries to achieve their goals.
• Consider joining international initiatives and partnerships that gather experts in the field and thought leaders to contribute to the development of AI for the benefit of humanity (such as Partnership On AI24 or City.AI).25
AI for Good
AI can drive both economic and social progress and help countries achieve national objectives like inclusive growth and development. But to achieve this, the technology must be developed in a way that is human-centred. African governments should consider the wider impacts of AI and implement policies accordingly. Responsible AI systems must be aligned with ethical values while empowering consumers.
There is a role for policy in ensuring that data and algorithms are used responsibly, and governments should support the collaborative development of codes of conduct by many stakeholders. Such codes should be cross-sector and based on shared, consensus principles. A recent report of the UK House of Lords suggests some principles that can form such a basis:
i. AI should be developed for the common good and benefit of humanity
ii. AI should operate on principles of intelligibility and fairness
iii. AI should not be used to diminish the data rights or privacy of individuals, families or communities
iv. all citizens should have the right to be educated to enable them to flourish mentally, emotionally and economically alongside artificial intelligence
v. the autonomous power to hurt, destroy or deceive human beings should never be vested in artificial intelligence.26
Further, governments should ensure that transparency, liability, accountability, justification, and redress for decisions are at the heart of the development and application of AI. These principles are essential to foster trust in AI and ensure its true potential can be realised by all.
This is in line with the African Union’s aspirations set in Agenda 2063 to build an Africa whose development is people driven, relying on the potential offered by people, especially its women and youth and caring for children.27
Artificial intelligence is an important opportunity for the continent of Africa. If governments can successfully navigate the challenges, AI can be a driver of growth, development, and democratisation. It has the potential to enhance productivity growth by expanding opportunities in key sectors for Africa’s development, including agriculture, healthcare, financial services, and government services.
By empowering them with access to high-quality digital tools, AI will equip workers, entrepreneurs, and businesses compete at a global level and be at the forefront of economic transformation.
However, the obstacles in the way require serious policy responses. AI will mean substantial adjustments for workers and business and opens new ethical questions that require thoughtful responses.
Labor and ethical questions are compounded by higher hurdles specific to Africa, stemming from gaps in connectivity, the readiness of education systems, and the availability of digital data.
Africa needs to take decisive steps to overcome its unique challenges, but if it can, it has the opportunity to catch up to those countries that have already taken steps to advance AI.
These efforts will not be easy, but the path forward is clear. Success will depend on the ability of governments to foster collaboration among all stakeholders — state and civil society, academia, industry, and national and international stakeholders.
If the challenges and opportunities of AI are embraced, Africa will reap the benefits of a vibrant AI ecosystem.
Download the entire, informative whitepaper to learn more about the opportunities and challenges for AI in Africa at this direct link: www.up.ac.za/media/shared/7/ZP_Files/ai-for-africa.zp165664.pdf
1If used effectively, AI can lead to important savings in public money. For example, in the UK, just the use of AI virtual agents across Government departments and the public sector is expected to save an estimated £4 billion a year. For more information, see techUK – Written evidence (AIC0203): data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/artificial-intelligence-committee/artificial-intelligence/written/70492.html
2Hila Mehr, Artificial Intelligence for Citizen Services and Government (ash.harvard.edu/files/ash/files/artificial_intelligence_for_citizen_services.pdf), Harvard Kennedy School, ASH Centre for Democratic Governance and Innovation, 2017
3Dina Ringlod, Citizens and Service Delivery: Assessing the Use of Social Accountability Approaches in Human Development Sectors, December 2011, The World Bank.
4The report also notes that the big challenge for these countries — and globally — is scarcity of AI developer talent. Governments need to re - think education for a future workplace redefined by AI and start building a healthy, collaborative, and open AI ecosystem to attract and retain competitive AI talent. According to a report by IDC, spending on cognitive and AI systems in the Middle East and Africa (MEA) region will reach USD 114.22 million by 2021, with the market expected to represent a compound annual growth rate of 32 percent for the 2017 to- 2021 period
5For more information on SANSTAR, see the Microsoft website (customers.microsoft.com/en-us/story/la-improves-sanitation-services-with-a-cloud-based-spatially-enabled-mobile-solution).
6For more information on Veritone, visit the company website at www.veritone.com/ai-solutions/government/
7For more information on NADER and NARRAI, see the Federal Ministry of Science and Technology website: scienceandtech.gov.ng/2018/08/01/fg-to-establish-two-new-agencies/
8For more information on Nigeria’s commitment to support AI, see the Federal Ministry of Communications website located at: www.commtech.gov.ng/news-and-media/daily-news-report/187-shittu-charges-private-sector-to-lead-in-artificial-intelligence.html
9Kenya OpenData. Kenya has also signed on to the Open Government Partnership: www.opengovpartnership.org/
10For more information on iHub, see the iHub website: ihub.co.ke/
11For more information on South Africa’s AI program, see Department of Trade and Industry website: www.dti.gov.za/industrial_development/fipt.jsp
12For more information on South Africa’s investments, see the government website: www.gov.za/speeches/minister-mmamoloko-kubayi-ngubane-netrepreneurs-rise-africa%20percentE2%20percent80%20percent99s-digital-lions-workshop-8-aug
13Philbeck, “Working Together to Connect the World by 2020: Reinforcing Connectivity Initiatives for Universal and Affordable Access, Broadband Commission, p.8
14“World Development Report 2016: Digital Dividends,” World Bank, 2016, p. 9
15J.M. Garcia and T. Kelly, “The Economic and Policy Implications of Infrastructure Sharing and Mutualisation in Africa,” World Bank Group, p.16
16For more information on New Broadband Pricing Data, see A4AI website: https://a4ai.org/broadband-pricing-data-2017/
17Tim Berners-Lee. 2009. Linked Data (www.w3.org/DesignIssues/LinkedData.html) (“Five Stars”).
18For additional recommendations targeted to the agricultural sector, see Global Open Data for Agriculture and Nutrition (GODAN), “A Global Data Ecosystem for Agriculture and Food,” August 2016, at https://www.godan.info/sites/default/files/documents/Godan_Global_Data_Ecosystem_Publication_lowres.pdf
19World Wide Web Foundation, Artificial Intelligence: Starting the policy dialogue in Africa, December 2017
20Accenture and Gordon Institute of Business Science, University of Pretoria, Artificial Intelligence: Is South Africa Ready?, p. 15
21A recent example of this is the Italian CINI National AI Lab, which gathered 600 AI researchers to push the country’s AI industry forward.
22Access to free, open, and anonymized, curated datasets is essential to train algorithms and support machine learning. It can also help improve accountability in AI development. Further, by encouraging stakeholders to use open government data in AI, the quality of data is improved.
23Brain drain” is a significant concern in this sector (as with other sectors in low- and middle-income countries), as local talent is systematically recruited by large U.S. and European companies working on AI, including Google, Facebook, and Apple. World Wide Web Foundation, Artificial Intelligence: Starting the policy dialogue in Africa, December 2017
24Partnership on AI: https://www.partnershiponai.org/partners/
25City.AI. For more information on the Tech in Ghana Conference, see event website at http://techinghanaconference.com/#overview
26Austin Clark, “Government ‘must play central role in AI development’” (www.govtechleaders.com/2018/04/16/government-must-play-central-role-in-ai-development/), GovTech Leaders,
16 April 2018.
27The Agenda 2063 (au.int/en/agenda2063) is a strategic framework for the socio-economic transformation of the continent over the next 50 years. Its builds on, and seeks to accelerate the implementation of past and existing continental initiatives for growth and sustainable development.
Excerpts from this paper are courtesy of Access Partnership and are published with their permission.
Access Partnership is a leading public policy firm that provides market access for technology. The company monitors and analyzes global trends for the risks and opportunities they create for technology businesses and identify strategies to mitigate those risks and drive the opportunities to our clients’ advantage.
The company’s team mixes policy and technical expertise to optimize outcomes for companies operating at the intersection of technology, data and connectivity.