Consumer law and artificial intelligence : challenges to the EU consumer law and policy stemming from the business’ use of artificial intelligence : final report of the ARTSY project

Abstract:

Potential regulation of use of artificial intelligence by business should minimize the risks for consumers and the society without impeding the possible benefits. To do so, we argue, the legal reaction should be grounded in an empirical analysis and proceed case-by-case, bottom-up, as a series of responses to concrete research questions.

The ambition of this report has been to commence and facilitate that process. We extensively document and evaluate the market practice of the corporate use of AI, map the scholarly debates about (consumer) law and artificial intelligence, and present a list of twenty five research questions which, in our opinion, require attention of regulators and academia.

The report is divided into four sections. The first explains our understanding of the concepts of “artificial intelligence” (a set of socio-technological practices enabled by machine learning and big data) and “consumer law” (various legal instruments concretizing the principles of the weaker party protection, non-discrimination, regulated autonomy and consumer privacy). The second section documents the ways in which the business uses artificial intelligence in seven sectors of the economy: finance and insurance, information services, energy and “smart solutions”, retail, autonomous vehicles, healthcare and legal services. For each analyzed sector we study the gains for the businesses stemming from the deployment of AI, the potential gains, but also challenges for consumers, as well as third party effects. In the third section, we repeat the analysis through the lens of four general “uses” of AI by businesses in various sectors: knowledge generation, automated decision making, advertising and other commercial practices and personal digital assistants. Finally, in the fourth section, we present the questions which we believe should be addressed in the next stage of the research. We cluster them into: normative questions about regulatory goals, technological and governance questions about regulatory means, and theoretical questions about concepts and preconceptions.