The role of governments as a provider of data for artificial intelligence

This project will survey the different ways in which governments are, could, and should, use data that they steward to support the development of AI. It will adopt or produce a suitable framework for understanding the types of data held by governments – such as administrative data, statutory registers, or routine monitoring data – and their potential to support the development of AI systems.

The first phase of the project will be to conduct a survey of data sharing by governments for the development of AI around the world, generating three illustrative case studies of current practices.

The second phase will examine how governments are making this data available to AI developers in well-governed ways – including how they are navigating issues around technology, culture, public attitudes, and fair financial models. This will create recommendations for governments who are trying to prioritise which data to make public, inform the discussion of how to navigate data governance challenges, and provide insight on the types of resources required to successfully govern the hosting of public data resources, particularly when these are sensitive in nature. The third phase will analyse state-of-the-art data sharing technologies and facilitators and who has access to them, and produce general guiding principles for the formulation of domestic rules in the interest of more transnational coherence.à

The intended impact is to increase the availability of publicly held data for AI grounded in the principles of human rights, inclusion, diversity, innovation, and economic growth by helping governments to prioritise their efforts and to reduce their concerns about the risks of sharing public data for AI by providing clear guidance, use cases and examples that demonstrate how it can be done safely and responsibly.