Data is powerful because it is used to tell stories – stories about people and the lives they lead. Data carries an appealing veneer of objectivity. But telling stories is never ethically neutral. Narratives always have embedded cultural values and ethical assumptions.

Data analytics and artificial intelligence are increasingly used to influence decisions about service delivery and access (in both the public and private sectors). In this talk I argue that the most vulnerable and marginalized citizens often rely on multiple government services, experience the greatest surveillance of their activities, and therefore have the most data produced about them. But conversely, they often have the least capacity to influence the narratives that are drawn with this data and the resulting policies.

To date, there has been significant discussion of data benefit sharing (ensuring the advantages/profits derived from the use of data are shared justly with the data providers and data subjects). I will discuss recent attempts, particularly in relation to indigenous data, to the move the debate from ‘benefit sharing’ towards ‘power sharing’. Power sharing requires a co-governance model where data subjects and communities have decision making capacity in relation to data governance and use.¹ I will describe recent initiatives in New Zealand, and argue in favour of transparency, accountability and co-governance models for public sector data.

If you would like to attend, please e-mail Jane Beinart at jane.beinart@ethox.ox.ac.uk.