As AI scales across sectors, robust data governance is needed to ensure shared rules serve the public interest and reflect diverse perspectives.
© Shutterstock/Gorodenkoff
Every time someone in a village clinic logs a patient record, a farmer checks crop prices on a phone, or a student submits homework online, they generate data. That data can train the next generation of artificial intelligence, guide public health decisions, or power services that didn't exist a decade ago. But the rules for who controls it and who benefits from it are still being written. Oftentimes, developing countries producing the data don’t find a seat at the negotiating table.
"We think of data like soil, not as oil, but like soil," says Isabel de Sola Criado of the UN Office for Digital and Emerging Technologies. Data governance, she explains, is like how societies use their soil: making sure it's healthy enough to grow something everyone can use.
A puzzle of rules
Yet this has been difficult. Different countries and regions built separate data rules that don't talk to each other. Information and the benefits it creates struggle to move across borders with ease.
But de Sola Criado sees change coming: "We are coming towards a convergent position," where governments, businesses and civil society are increasingly open to sharing more data for development and for AI — as long as there are shared principles and safeguards everyone can agree on.
Small steps, not one rulebook
Nobody expects an immediate solution. The goal, de Sola Criado says, is "not trying to solve the entire picture, but going with what works".
It’s with small, workable steps towards a shared way of managing data for development, rather than one rulebook handed down from outside.
UNCTAD's role in writing the rules
That incremental approach is exactly what UN Trade and Development (UNCTAD) is helping to build.
As secretariat of the UN Commission on Science and Technology for Development since 1993, UNCTAD now supports a dedicated working group on data governance.
It brings together 27 government representatives and 27 non-government members, from academia, business and civil society, to hammer out common principles for how data should move and who gets a say in the rules.
The group's findings will feed into a report to the UN General Assembly.
Why this matters now
The stakes keep rising. As artificial intelligence spreads into health care, education, agriculture and finance, the data feeding those systems needs governance that keeps pace, before the rules are set entirely by the countries and companies that already dominate the field.
That conversation continues this month, as global leaders gather in Geneva for the UN's first Global Dialogue on AI Governance and the AI for Good Summit.