The European Union has committed to the principles of responsible and trustworthy artificial intelligence (AI) by requiring mandatory risk assessments for AI systems prior to their deployment. Prior to its introduction, AI-based systems were subject to different jurisdictions under different aspects (e.g. product liability), whereas the AI Act provides a comprehensive and uniform framework legislation for AI systems in the EU. Its introduction has shifted the debate about 'good' AI from the discussion about ethical guidelines and voluntary commitments to the binding nature of the law. At the same time, we are in a transition phase, as the standardization and national implementation of the EU AI law are not yet fully completed. So far, there is little empirical evidence on how organizations, especially small and medium-sized enterprises (SMEs), are preparing for these new requirements.

Against this background, this exploratory project examines how SMEs in Ostwestfalen-Lippe (OWL) - one of Germany's leading innovation locations - are reacting to the EU AI Act. With the help of expert interviews with managing directors and AI managers, we analyze to what extent SMEs are aware of the law, what concrete measures they have taken or are planning and what challenges they identify in its implementation. These results are embedded in the context of larger debates about responsibility gaps in technology ethics. In particular, the focus will be on challenges such as distributed action (“many hands”), complex technical systems (“many things”) and specific AI issues.

Cooperation partners:

This project is being carried out in cooperation with the Association of German Engineers (VDI) and the Technical University of Munich (Prof. Dr. Sebastian Pfotenhauer).