UNESCO and the World Bank recently hosted a workshop in Conakry, Guinea, bringing together African policymakers and industry leaders to discuss sustainable artificial intelligence (AI) development. The event aimed to explore strategies for integrating environmental considerations into national AI strategies as AI adoption increases across the continent.
Workshop Focus and Participants
The workshop, held November 11, included policymakers, regulators, and industry leaders alongside experts from UNESCO and the World Bank. Discussions revolved around “Green AI,” acknowledging AI’s potential to support development while also recognizing its resource intensity regarding energy, water, and land usage. Participants analyzed the AI value chain through an environmental perspective, examining trade-offs throughout the lifecycle of AI systems.
Approaches to Green AI
Experts presented two distinct approaches to achieving sustainability in AI. “Greening AI” focuses on improving the efficiency of AI systems themselves, while “Greening with AI” explores using AI to support environmental and sustainability goals.
Resource Efficiency Strategies
UNESCO presented findings from its report, Smarter, Smaller, Stronger: Resource-Efficient Generative AI and the Future of Digital Transformation. The report detailed technical methods for reducing AI’s resource demands, including model compression techniques like quantization, which can reduce energy consumption by up to 44%. Shorter prompts and outputs can reduce energy use by more than 50%. The report also found that task-specific small language models can achieve up to 90% energy savings compared to large language models while maintaining accuracy.
The Role of Policy
UNESCO and the World Bank emphasized the importance of policy levers in promoting sustainable AI. National digital and AI policies influence the development of data centers and telecommunications infrastructure. Sector-specific policies related to water, energy, mineral resources, and environmental management also play a crucial role in shaping the sustainability of AI ecosystems.
Case Studies in Policy Frameworks
The World Bank presented case studies from Nigeria, Singapore, and Chile to illustrate how different policy frameworks can support green AI across various contexts.
Facilitation and Regional Efforts
The workshop was facilitated by Leona Verdadero, a program specialist in digital policies and digital transformation at UNESCO, and Hawa Siga Diankon, a digital development project associate for West and Central Africa at the World Bank. The session was part of the seventh Transform Africa Summit and contributed to broader regional efforts to develop an AI-ready digital future. UNESCO stated its commitment to continuing support for African countries in building inclusive and sustainable AI ecosystems and strengthening capacity for green digital transformation.
