Skip to main content

Day 3 - Successfully harnessing AI in Africa

December 2 - December 4 2024 I WP3458

A,Shot,Of,A,Smiling,Female,African,Farmer,Using,A

Session 5: Breakout groups

Day 3 consisted of focused discussions in small groups to consider in very practical terms how to “turbocharge” existing AI projects and initiatives in Africa including by leveraging Google’s technologies, especially in four areas: skilling for AI; AI-driven scientific research and innovation; data and data governance; and AI ethics, safety, and trust. The recommendations were as follows:

AI Skilling

Group discussions on AI skilling emphasized the urgent need to scale AI education across Africa through targeted training programs involving governments, tech companies, and schools and universities. Participants identified three critical areas for action: mainstreaming AI into education systems at all levels, creating inclusive skilling programs for underrepresented groups, and enhancing workforce readiness through industry-driven training.

Among the recommendations were (1) mainstreaming AI learning into primary, secondary, and tertiary education curricula, with tailored programs for rural populations, women, and persons with disabilities; (2) developing an AI skilling blueprint to inform strategic decision-making about investment in skills; (3) scaling existing initiatives like Google’s AI literacy courses and Google’s Hustle Academy for SMEs; and (4) building on existing initiatives like the AI for Science Master’s program at the African Institute for Mathematical Sciences, AI4D responsible AI labs at African universities and African Union-led upskilling initiative. Participants also recommended integrating AI skilling into government leadership programs to ensure AI in policymaking and public service innovation. Specific next steps include promoting AI skilling models at the upcoming Global AI Summit in Rwanda and launching industry-academic collaborations for applied AI research.

Additionally, the discussion underscored the need for harmonized AI education policies across regional economic blocs like the Intergovernmental Authority on Development (IGAD) and Southern African Development Community (SADC).

“Participants urged tech companies like Google to open their AI safety tools and offer capacity-building programs tailored to African regulatory contexts.”

AI scientific research in Africa

Discussions focused on mainstreaming AI into Africa’s scientific and innovation ecosystems, with an emphasis on addressing the Sustainable Development Goals (SDGs).

Participants in particular urged shifting from isolated pilot projects to systematic, large-scale applications aimed at mainstreaming AI in healthcare, agriculture, and climate resilience. Proposed actions included (1) embedding UK PhD students in African scientific labs in various disciplines; (2) creating joint funding streams for early-career researchers; and (3) launching challenge-based innovation programs led by DeepMind scientists. There were also calls for expanding research centres dedicated to AI in Africa and building robust partnerships with global scientific institutions.

There was consensus on the need to link scientific data collection with commercial viability. For example, AI adoption and research could be incentivized through innovation hubs focused on commercializing AI-based solutions. The conversation also explored scaling AI tools like Jacaranda Health’s maternal care system and piloting AI-powered digital rehabilitation models.

Frameworks on data governance for Africa

The discussions highlighted that unlocking Africa’s AI potential depends on advancing balanced and harmonized data governance frameworks, cross border data transfer, promoting open data, and creating data-sharing mechanisms. There is much experimentation and debate on innovative approaches to data governance for equitable data sharing which could be applied to help unlock data in a responsible way.

Recommendations included (1) explore ways to grow the FCDO, IDRC and Google-supported Lacuna Fund to provide open datasets in critical fields like climate, agriculture, and sign language; (2) launching an advocacy campaign for open government data; and (3) building capacity within government agencies to recognize data’s economic value. Kenya’s White Paper on data as an economic asset, aimed at enabling policymakers learn about the ways in which AI uses and requires data, was cited as a model for similar thinking among African governments. Open cross-border data transfers were seen as essential to building robust and unbiased AI models.

In addition, attention was drawn to the need for data for inclusive AI development. To enhance data accessibility, the participants proposed developing collaborative datasets for underrepresented segments, such as sign language and indigenous languages. Establishing a continental data protection authority was recommended to coordinate policy harmonization under the African Union’s digital protocol.

AI safety and ethics for Africa

The session on AI ethics and safety stressed the need for a continental AI ethics and safety framework aligned with global standards. The participants noted that existing efforts, including UNESCO’s AI ethics framework, should be connected to African-led initiatives for greater regional representation and developing clearer conceptualization of AI ethics in an African context.

Proposals included (1) considering holding a Global AI Safety Summit in Africa to establish guiding principles on AI ethics and safety in an African context; (2) embedding AI ethics experts in projects across sectors, particularly in sensitive areas like healthcare and public services; (3) expanding responsible AI assessments and corporate governance models for SMEs; and (4) exploring collaboration between Google and African researchers on benchmarking or quality assuring AI tools in a local context.

There was a call to ensure African values are represented in global AI safety standards through structured participation in global bodies like the G7. Participants urged tech companies like Google to open their AI safety tools and offer capacity-building programs tailored to African regulatory contexts.

Session 6: Partnerships: Working together to unlock the AI opportunity in Africa

Participants concluded that partnerships across sectors, borders, and disciplines are essential to unlock Africa’s AI potential. The discussion emphasized the role of targeted investments, open data initiatives, and joint capacity-building programs.

A development leader outlined how his organization supports AI policy innovation hubs in 12 African countries, helping them build policy frameworks, data-sharing agreements, and language technology initiatives. USAID’s digital programs promote inclusive AI ecosystems by combining digital infrastructure investments with AI-specific training programs. Their focus on last-mile connectivity ensures that AI-powered innovations reach underserved communities. Another expert discussed the role of trade policies in facilitating AI data flows, noting that Africa’s trade agreements should drive at open data-sharing standards to avoid restrictive data localization practices.

Key recommendations from the session on partnerships for AI included:

  • Leveraging existing, well-working partnerships and initiatives that work to promote AI development: African governments and development organizations can readily promote AI development by partnering with companies like Google and leveraging existing initiatives, like those led by Google-DeepMind to promote AI among Africa’s scientific communities, Google Hustle Academy to scale AI use among SMEs, and the Lacuna Fund that promotes data for AI.
  • Building national AI skilling programs: Countries should adopt national AI skilling blueprints integrating technical, ethical, and policy training. Partnerships with universities, tech firms, and development agencies can expand these programs’ reach and impact.
  • Promoting cross-border data frameworks: African governments should create data-sharing standards that allow for secure, equitable cross-border data flows. Establishing a continental data protection body would reinforce compliance and harmonization.
  • Investing in AI innovation funds in Africa: Public and private funders should establish targeted funds supporting African AI startups, research centers, and tech-driven businesses. Blended financing models can accelerate commercial-scale AI innovation.
  • Safety and Ethics: Developing an African-led framework for AI safety and ethics.

Previous

Day 2 – Successfully harnessing AI in Africa

Next

Conclusion: Seizing Africa’s AI opportunity through collaboration

Want to find out more?

Sign up to our newsletter