Moreover, only 51.5% of Sub-Saharan Africa had access to electricity as of 2022, with an average of 87 blackouts annually, further impacting AI development.

To enhance AI safety, African researchers should frame AI issues within the context of local values and engage in collaborative funding efforts.

African countries are increasingly making their mark in the global AI landscape, with notable growth in developers, particularly a 45% increase in Rwanda and Nigeria between 2022 and 2023.

To capitalize on open-access AI and enhance safety, African developers should prioritize international collaborations, local safety networks, and standardized evaluation practices.

Access to critical AI infrastructure, like GPUs, remains a significant barrier in Africa, where only 1% of Zindi Africa data scientists have on-premises access, compounded by high costs relative to GDP per capita.

China's significant investment in African AI technologies raises concerns about the potential influence on African collaboration with Western partners, particularly in the context of the Belt and Road Initiative.

Establishing safety research networks, similar to the European Network for AI Safety, is essential for pooling resources and coordinating AI safety initiatives across the continent.

Additionally, leading AI nations are adopting nationalistic policies that could alienate developing countries, including those in Africa, from vital AI resources and collaborations.

However, African AI safety researchers often face exclusion from global research networks due to financial constraints and visa issues, with AI safety research comprising only 2% of total AI research from 2017 to 2022.

Despite the growth in the sector, challenges such as dependency on foreign model-sharers and systemic developmental issues may hinder progress in AI safety research.

The U.S. Bureau of Industry and Security's new licensing requirements complicate access to advanced AI models for African nations, as none are considered low-risk destinations.

This control over AI models and computing resources by higher-income countries may limit African developers' access to the advantages of open-access AI.