Encouraging the Development of Ethically and Societally ‘Desirable’ Technologies Through Innovation Governance

Society is currently undergoing massive transformations linked to the so-called Digital Revolution. Big Data, Artificial Intelligence, Machine Learning and associated innovations are profoundly impacting our social, economic, political and cultural life. Technologies are, however, not neutral tools (Hubig et al. 2001). They are not developed in a societal or political vacuum; norms, values and interests are part of their substance and effect. Technologies such as Machine Learning are associated with various ethical challenges, including risks to justice, privacy and human autonomy. Some technological innovations may also negatively influence democratic processes. Combining ethics of technology, sociology and political science, we argue that it is essential to analyze technologies’ often intransparent development processes as well as their contexts of development in order to understand innovations’ underlying political conditions, their normativity, and impact. This opens up a window of opportunity for practicing (a degree of) democratic control over the process of technological innovation and its outcome. Our paper draws on the concept of governance as both a useful analytical tool to grasp the multi-level complexity of socio-technical systems, and a guiding normative concept. We propose an extended concept beyond ‘active’ governance by state and non-state actors, which also includes passive governance mechanisms. By regarding Governance by Infrastructure and Governance by Algorithms (Musiani 2013; DeNardis/Musiani 2016), the normative and formative powers of technology itself become subjects of analysis. Such passive governance through technology, framework conditions, and infrastructure amplifies the ‘democratic deficit’ of technology innovation processes. Our contribution ultimately poses the normative question: How should governance processes be designed to encourage the development of ethically and societally ‘desirable’ technologies? By exemplarily assessing the configuration and influence of existing innovation governance mechanisms, we seek to identify strengths, weaknesses and potential best practices. Based on an ethical analysis, we then develop feasible and concrete ‘points to consider’ for the design of ethically acceptable innovation governance. These may be relevant e.g. for engineers and IT-developers, research funding bodies and political decision-makers. Our paper will present qualitative case studies of national (German) and supranational (European) governance mechanisms of technological innovations. Based on literature analyses, expert interviews and participatory observation, we examine the relevant actors, technologies, infrastructures, interests and power relations. For instance, regarding autonomous driving, multi-level actors’ ever-evolving attempts to regulate and control innovation include (inter-)national laws and technical standards, targeted research funding, and expert bodies such as the German Ethics Commission for Autonomous and Inter-Connected Driving. Passive determinants such as infrastructure conditions (traffic lights, network coverage etc.) and societal expectations are also influential. Other technological innovations such as social bots, deep fakes, and specific algorithms which may be creating ‘filter bubbles’ or ‘echo chambers’ might pose more fundamental challenges to democratic values and processes, including agenda-setting, elections and decision-making. Hence, the governance of these innovations is of great political relevance.