Dinner Speech - Silja Vöneky

How Should We Regulate Artificial Intelligence? A Starting Point 

One major challenge of the 21st century to humankind is the digital transformation including the widespread usage of Artificial Intelligence (A.I.). There is hardly any day without news about the disruptive force of A.I. – both good and bad. Some warn that A.I. could be “the worst event in the history of our civilization”. High-tech surveillance in some States seems to create a new automated authoritarianism. Other stress the chances of A.I., diagnosing – for instance – cancer, or helping humans as social robots and autonomous cars. But as A.I. is so disruptive the call for its regulation is widespread, including international treaties banning for instance “killer robots”. Nevertheless, it is unclear how and to which extent we should regulate A.I. I will exemplify what has been done so far as part of “bottom up” governance and “top down” regulation approaches, and I will show how new codes, as the OECD Principles on A.I., international treaties and national laws, can be part of a solution to limit risks of A.I. without undermining science, and innovation. Besides, I will propose ways in coherence with human rights for the challenges posed by A.I. that lay ahead of us. In the end, I will argue why European approaches to responsible A.I. might and should inspire rules and norms in other States and at the international level.