Challenges of the Knowledge Society (May 2024)

AUTOMATED DECISION-MAKING: IS THE EU CONSUMER LAW FIT FOR THE EMERGING TECHNOLOGY?

  • Monica CALU

Journal volume & issue
Vol. 17, no. 1
pp. 148 – 153

Abstract

Read online

The concept of „automated decision making” (ADM) is currently a key vector reshaping industries in the era of technology-based decision making, and the use of artificial intelligence (AI) in this process is a hot topic worldwide with the adoption of generative AI tools like ChatGPT. ADM enables companies to use technology for tasks that were not long ago the domain of human judgment. AI is essential in financial services, such as credit worthiness assessment, automated investment advice (often called robo-advice) or insurance. The regulation of activities related to ADM is contained in various acts and regulations, including the Sale of Goods Directive 2019/771, the Digital Contracts Directives 2019/770 and the EU General Regulation on Data Protection. However, these directives do not contain special rules for new technologies such as Artificial Intelligence (AI). Although recently, the EU has moved forward with regulations on product liability laws affecting digital products and services, with the EU Data Act, the EU Cyber Resilience Act and the highly publicized EU AI Act - regulations with a transition period of two years for compliance, the speed with which in practice the ADM tools are developed and trained make the legislative efforts slow and incomprehensible. In light of this issue, this article explores the relationship between AI, ADM and discusses whether existing EU consumer law is equipped to deal with situations where AI systems are either used for internal purposes by companies or offered to consumers as the main subject of the contract. The paper will reveal a number of gaps in current EU consumer law and briefly discuss future legislation. This article also reviews the efficiency of key EU consumer law directives for the conclusion of contracts with consumers through the use of AI, with recommendations for clarifications and additions to improve the use of ADM.

Keywords