UN data governance discussions enriched with Alliance input

The UN Commission on Science and Technology for Development (CSTD) has learnt out African progress in harnessing data for development, in an event held on 6 November in Geneva. On behalf of the African Alliance for Access to Data, RIA Executive Director, Dr Pria Chetty, highlighted ongoing African initiatives to both protect and avail data, including the AU Data Policy Framework, and the ACHPR African Guidelines on Access to Data.
Her presentation explained how African initiatives relate to the context where: “AI depends on data across a lifecycle: Data must be generated or collected, documented, accessed, processed, used to train or evaluate systems and reused over time. Governance decisions at each stage shape the outcomes.”
The public-interest question about data for innovation, said Dr Chetty, is “whose data are collected, whose knowledge is represented, who can access the data and who benefits from AI applications?”.
She noted that valuable data may be controlled by large digital platforms and other private entities, and that while “Language resources may exist in books, community knowledge, broadcast recordings, archives or research collections”, these remain undigitised, poorly stored or difficult to access.”
Her presentation further unpacked the impact of asymmetries in data representation, value distribution, design, and ownership across the technological value chain. “Equitable data governance is not just about who can generate and use data, but rather, who shapes its rules and benefits most from its value,” she said.
Equitable stewardship for data should address four questions, according to Dr Chetty:
1. Representation: Which languages and communities are present in the datasets?
2. Access: Who can access and use the datasets and under what conditions?
3. Control: Who decides how language resources are collected, stored, licensed and reused?
4. Benefits: How do language communities and local researchers benefit from the systems built using their linguistic resources?
Concretely, she urged CSTD to help address these issues by:
• Bringing data governance, AI policy, digital infrastructure and development institutions into the same conversation.
• Catalysing multi-stakeholder dialogue between governments, researchers, private data holders and human rights institutions.
• Supporting international cooperation on access to data relevant to cross-border public-interest research.
CSTD should also share practical institutional models for regulatory coordination, impact assessment, oversight and accountability, she recommended.
