Continental AI strategy for 2025-30 to feed into UN summit

The African Union’s Executive Council has adopted a Continental Artificial Intelligence (AI) Strategy and African Digital Compact proposed to run from 2025 to 2030 which will be the region’s contribution to the United Nations Summit of the Future to be held next month.

In a statement on 9 August, the African Union announced its Executive Council’s decision that came after African ICT and communications ministers had previously unanimously endorsed the strategy.

The African Union Commission has said 130 African ministers and experts convened virtually for the Second Extraordinary Session of the Specialised Technical Committee on Communication and ICT in June to guide the continent on changes that are being fuelled by AI technology and applications.

The adopted Continental AI Strategy wants to harness AI to meet development aspirations while, at the same time, promoting ethical use, minimising potential risks, and leveraging opportunities. It includes a proposal for an AI Ethics Board that can review research applications in the field.

ICT ministers also mandated the African Union to organise a Continental African AI Summit to foster collaboration, knowledge exchange and strategic planning among stakeholders across the continent.

African Digital Compact

The African Union Commission said in a statement that the strategy and the Digital Compact were developed through consultations with a wide array of stakeholders and constitute Africa’s contribution to the Global Digital Compact and the United Nations Summit of the Future in September 2024.

The statement said the strategy covers education and skills, climate change, health, agriculture, infrastructure, peace and security and good governance, among others.

“By investing in African youth, innovators, computer scientists, data experts and AI researchers, the framework paves the way for Africa’s success in the global AI arena,” according to the statement.

It said the strategy was informed by previous initiatives and AU strategic frameworks and benefited from the guidance and support of a task force composed of experts as well as the outcome of virtual multi-stakeholder consultations held in April 2024.

The statement, explaining the strategy, says there are a number of “action areas” that include promoting research and innovation in AI, through partnerships between academia and the private and public sectors.

It said there is a need to invest in African-led research on the short, medium and long-term risks of AI to African societies, including through research funding, fellowships and the establishment of research chairs.

The potential applications and risks of AI in education demand the modernisation of curriculums and this “encourages a shift away from educating for specific jobs towards acquiring skills that allow graduates to adapt to evolving tasks arising from technological change”, the strategy said.

“Africa should also build AI skills in higher education. This requires integrating AI into computer science and mathematics education and establishing advanced research in various AI domains,” according to the strategy.

It noted that a designated centre of excellence could help to conduct research and build a knowledge base in the field.

Challenges

According to the strategy, Africa faces a number of challenges and barriers limiting AI uptake, notably gaps in internet usage, a lack of computing platforms, limited data availability for training AI models, and a scarce supply of AI skills, which inhibit the use of AI for social and economic development.

Although internet usage is increasing, it is still not extensive enough to support AI development due to a number of factors, including affordability.

The lack of data is a barrier to developing solutions that use language understanding or speech recognition in different African languages. Limited data infrastructure, such as data centres, storage facilities and high-performance computers, is also critical, the strategy added.

According to the strategy to date, five countries, Benin, Egypt, Mauritius, Rwanda and Senegal, have developed standalone AI strategies, while others, such as Kenya, South Africa and Uganda, are addressing AI in combination with other emerging technologies, such as blockchain or in the context of the Fourth Industrial Revolution.

Women and AI

The upcoming Summit of the Future presents a critical opportunity to shape the trajectory of AI in Africa by ensuring that it aligns with the continent’s unique needs, values and aspirations, said Simisani Ndaba, a teaching assistant in the department of computer science at the University of Botswana, who is also the founder and co-organiser of R-Ladies Gaborone, a chapter of R-Ladies Global, a worldwide organisation that promotes gender diversity by utilising R, a free software environment for statistical computing and graphics.

She told University World News the most important thing the AU should keep in mind is localised solutions: “The summit could emphasise the importance of developing AI technologies tailored to Africa’s specific challenges and opportunities, such as [the fields of] agriculture, healthcare, education, and infrastructure.”

Ndaba said African governments must ensure that women and young girls are not left out in AI strategies. This can include running national campaigns to raise awareness about the importance of gender diversity in Al.

She said scholarships, mentorship programmes, and targeted campaigns can encourage more women to pursue careers in AI and related fields.

Ndaba said there is a need to define measurable goals, such as increasing female enrolment in AI courses by a certain percentage or achieving specific quotas for women in AI jobs.

“There is a need to set up channels for feedback from women and girls involved in AI initiatives, allowing the strategy to be refined and adjusted as needed. Be prepared to adapt the strategy to address new challenges, such as changes in AI technology or shifts in the job market, ensuring that women and girls continue to benefit from AI advancements,” she added.

What lecturers say

A lecturer at Zimbabwe’s Manicaland State University of Applied Sciences department of agricultural economics and development, Fortune Mijeri, told University World News that African universities can benefit from AI by integrating it into curricula, conducting research and collaborating with industry.

“However, older lecturers are resisting, due to fear of replacement, lack of understanding, or scepticism about AI's benefits,” said Mijeri.

To address this, he added, there is a need to provide AI training and support for faculty staff, encourage collaboration and knowledge sharing of information and showcase successful AI integration examples from developed countries.

“Universities can also develop interdisciplinary AI programmes and training, foster international partnerships between developed and developing countries and promote diversity and inclusivity in AI education and research.

“Africa is still reluctant to adopt AI. So, personally, I feel it is high time that we took a holistic approach and tried to change the way curricula is imparted to students. Also, research projects now need to be reassessed as AI has influenced how research is done,” he added.

Tumelo Bethuelson, who is employed as a data science manager at a financial institution in Botswana, said that, for the most part, the strategy is solid.

He said that, since the strategy is more of an AI governance proposal, it is too early to pinpoint any deficiencies.

“Africa tends to have its own unique problems requiring specialised solutions and I believe the whole thing will be a learning experience for both the AU and its member states,” he said.

“As stated above, it will be a learning experience. It should be left to individual member states to define and chart their own way forward as they each understand their individual peculiarities. They have to do this, of course, without deviating too much from the guidance of this strategy. Usually implementation is a different ball game from planning.”

Bethuelson said universities need to offer strategy-aligned courses and research and development should be core to most technology-oriented institutions.