Continental Artificial Intelligence Strategy (African Union)

Artificial intelligence (AI) has great potential for Africa’s socio-economic transformation and cultural renaissance. It will contribute to the attainment of the African Union Agenda 2063 and Sustainable Development Goals. It presents huge opportunities for inclusive development - creating jobs, improving the delivery of public services, advancing agriculture, education and health, promoting access to information, and protecting the environment and natural resources, among others. However, these benefits are accompanied by AI risks and harms, ranging from bias due to the way data were collected and trained on AI applications, potential discrimination of women and vulnerable persons (migrants, children, persons with disabilities), to job displacement, effect on indigenous knowledge and liability issues due to AI overtaking certain human operations. The risks are being deepened by Generative AI, which include disinformation, infringement of data privacy, surveillance, and copyright violations. 

AI systems may perpetuate or amplify biases contained in datasets that they are trained on because, more often, data are not equitably sourced - data are usually sourced from developed countries and from non-diverse and inclusive developers' teams. Besides, AI systems may not yet be fully able to explain their decision-making. AI also present concerns for the overall protection and promotion of human rights and poses safety and security issues in civil and military settings. This ranges from cyber threats to AI applications in disinformation for manipulation of political systems and societies to Generative AI-produced and distributed information, which can be fully automated through untraceable deepfakes. 

There is also an AI divide between and within countries and globally. These and other emerging risks exacerbated by advanced AI systems making them more complex to address, require the adoption and implementation of principles, frameworks and adaptable regulations on how to design, train and operate AI systems in a trusted and ethical manner, respecting the rights, culture and values of the African population.

To realise AI's positive and transformative potential for African development and mitigate potential risks, it is imperative to build the necessary capabilities. These range from sustained investment in infrastructure (reliable and efficient electricity, broadband connectivity, data infrastructure like data centre and cloud, and computing power) to huge sets of quality data, education and skills in AI, media and information literacy (MIL), research and innovation. Africa also needs to build a vibrant and inclusive AI start-up enterprise ecosystem that develops and implements social and economic applications and systems. Regional coordination and international partnerships are also needed to help Africa build the capacity to leverage AI in a manner that promotes social and economic development and preserves its peace and security. Finally, African AI systems require concerted investment by governments and the private sector in the region. The Continental AI Strategy proposes a people-centric, development-oriented and inclusive approach around five focus areas and fifteen policy recommendations. 

The five focus areas are:

• Harnessing AI's benefits for African people, institutions, the private sector and countries, in line with Agenda 2063 (i.e., improving people's livelihoods, leaving no one and no place behind, with a focus on AI applications in agriculture, education, 4 healthcare, public service delivery, climate change, and peace and security) and promoting the competitiveness of the African private sector.
• Addressing the risks associated with the increasing use of AI, with attention to governance, inclusion and diversity, human rights, gender equality, dignity, safety, peace and security, information integrity, and sustainable environment and ecosystem, considering African contexts, cultures, and values.
• Accelerating AU Member States' capabilities in infrastructure (energy, broadband, computing capabilities, data centres, cloud, IoT), AI talent and skills, datasets, innovation and research that underpin AI development,
• Fostering regional & international cooperation and partnerships to develop national and regional AI capabilities and advance Africa’s position on a global stage.
• Stimulating public and private investment in AI at national and regional levels.

As outlined in below diagram, the Strategy proposes fifteen action areas: 

• The first action area provides for the establishment of an appropriate AI governance system and regulations at regional and national levels. 

• The second action focuses on promoting the adoption of AI in the public sector, with a view to delivering efficient services to citizens, businesses and others. 

• The third area of action aims to accelerate the adoption of AI in the core sectors outlined in Agenda 2063 and the Sustainable Development Goals (SDGs). Sectors with high social and economic value include agriculture, education, health, culture, climate change and adaptation. 

• The fourth area of action will be the adoption of AI by the private sector, including small and medium enterprises. 

• The fifth area of action focuses on creating an enabling environment for a vibrant and inclusive AI startup ecosystem. 

• The sixth area of action aims to ensure the availability of high-quality and diverse datasets for AI. Data in an open format or through regulatory sandboxes is critical for the development of social and economic solutions. This action area also focuses on building the underlying infrastructure for AI, like computing platforms such as highperformance computing, data centres and cloud services. 

• The seventh area of action focuses on information integrity, media and information literacy. 

• The eighth action area focuses on promoting AI skills and talent in schools, colleges, workplaces and among the population, including the provision of upskilling and reskilling programmes and requirements for jobs at risk. 

• The ninth action area promotes research and innovation in AI through partnerships between academia and the private and public sectors. This action line promotes the development of challenge-driven AI research in priority areas through collaboration between the research community and the public and private sectors. 

• The tenth area of action provides for the adoption and implementation of ethical principles for AI that respect:

a. Human rights and dignity within the framework of regional and international human rights law. 

b. Diversity, inclusivity and African culture and values (e.g. the inclusion of women as well as vulnerable persons including migrants and people with disabilities and values such as Ubuntu, which respects collective community over individuality). 

c. Promotion of the protection of children in the face of AI as Africa is fast becoming a young continent. 

d. Responsible and unbiased use of AI, respecting intellectual property rights, and addressing potential societal and legal implications

• The eleventh area of action provides for the adoption and implementation of technical standards to ensure the safety and security of AI systems across the Continent. 

• The twelfth action area aims at accelerating public and private investment in AI in Africa. 

• The thirteenth action area focuses on promoting regional cooperation and coordination with the participation of relevant stakeholders from academia, civil society, media, governments, and the private sector to maximise the benefits and minimise the risks of AI in Africa. 

• The fourteenth area of action provides for accelerated African participation in global AI governance. 

• The fifteenth area of action promotes AI-related partnerships between Africa and the rest of the world, with a view to mobilising financial and technical resources for AI development in Africa.

The complete strategy can be found here.