Towards AI Governance in DAX40: A Typology of Organizational Guidelines for Self-Regulation

In this article, we combine the discourse on the ethical challenges of using Artificial Intelligence (AI) with research perspectives on overarching AI governance. Although AI govenance is not yet institutionalized, an increasing number of organizations are formulating their own guidelines for the responsible use of AI. These self-regulatory approaches serve as guidance for customers and employees. At the same time, they serve to align organizational processes and control mechanisms. Little is known about the differences between self-regulatory approaches and how organizations anticipate the future direction of AI governance. Using DAX40 companies as an example, we examine how organizations design AI guidelines for self-regulation and which criteria of AI ethics they take into account. Based on a systematic search strategy and qualitative content analysis, we identify three different types of self-regulation: (1) non-codified self-regulation, (2) symbolic-technical self-regulation, and (3) comprehensive socio-technical self-regulation.