IGF 2020 Pre-Event #61 Data for Good AI: How to create data governance model that fosters socially beneficial AI?

Roundtable discussion with presentations. The session will focus on discussing the issues of digital economy, especially data-driven innovations. Data sharing is still not prevalent enough and in many fields like healthcare, education or smart city the lack of access to quality data prevents from training AI models and creating novel products and services. The few tech companies that do have data often keep them behind IP laws and walled gardens, thus preventing developing countries, SMEs and youth entrepreneurs from fair competition. As the COVID-19 crisis has shown, there is a clear need for data-driven innovations in public interest, especially in provision of public services (i.a. telemedical solutions, e-learning). We will explore the following questions that expand the main one captured in the title of the session: • What are the priority areas for data-driven public good / public interest? • Who holds valuable commercial data sets and how to create a viable alternative governance model for them? • How to foster training of socially beneficial AI? The outcomes will allow to understand how the new governance models for data can allow for fully utilizing the value of information and therefore boost the digital economy across countries and industries.

The roundtable will begin with presentations by:

• Jan J. Zygmuntowski - Kozminski University, Instrat
• Dimitar Dimitrov - Wikimedia Belgium
• Stefaan G. Verhulst - NYU Gov Lab
• Alexandra Przegalinska - Kozminski University, MIT
• Nicolas Moës - European Horizons, Future Society
• Michael Donaldson - Commissioner for Digital Innovation, Barcelona City