From a ‘race to AI’ to a ‘race to AI regulation’: regulatory competition for artificial intelligence 论文

摘要

Against a background of global competition to seize the opportunities promised by Artificial Intelligence (AI), many countries and regions are explicitly taking part in a ‘race to AI’. Yet the increased visibility of the technology’s risks has led to ever-louder calls for regulators to look beyond the benefits, and also secure appropriate regulation to ensure AI that is ‘trustworthy’ – i.e. legal, ethical and robust. Besides minimising risks, such regulation could facilitate AI’s uptake, boost legal certainty, and hence also contribute to advancing countries’ position in the race. Consequently, this paper argues that the ‘race to AI’ also brings forth a ‘race to AI regulation’. After discussing the regulatory toolbox for AI and some of the challenges that regulators face when making use thereof, this paper assesses to which extent regulatory competition for AI – or its counterpart, regulatory convergence – is a possibility, a reality and a desirability.