Between Regulation and Promotion: Coping with the Challenge of Artificial Intelligence in Germany and the UK

While Artificial Intelligence (AI) has been around for decades, things have changed drastically after the publication of ChatGPT in November 2022: With the new application gaining more than a million users in the first five days, the new variant of ”generative AI“ (GenAI) became the fastest growing technology in history. Given that GenAI is purely knowledge-based (i.e. required e.g. no special natural resources), success in this field – that is supposedly going to revolutionise the economy, if not the world – was in principle within reach for many advanced countries. As that this was likely a ”winner-takes-all“ technology, a race for domination started, while at the same time the necessity for regulation of this new technology (which some of its most prominent pioneers fear could be so dangerous as to cause human extinction) was clearly recognized. This meant a double challenge for policy makers who had to aim both for promotion of GenAI in their economy and for effective attempts at regulation. This paper describes and analyses how they coped with this challenge by comparing the developments in two economically similarly developed but politically-institutionally quite different countries, namely the UK and Germany. After a descriptive part outlining the timelines, decisions, and measures for each country, the questions addressed in the comparative analysis include: * Which balance did policymakers strike between regulation and promotion? * Which priorities did they pick, given the multiplicity of potential topics (which range from technological innovation and market support to regulatory concerns about employment, security and human rights)? * Where did policymakers put the balance between public and private sector roles – is the former merely to act as a facilitator while the leading role is played by the latter, or is the state itself to assume the leading role? * Which instruments and incentives did they choose to implement their choices – a focus more on funding and subsidies (positive incentives) or on legal rules aiming to enforce and forbid certain developments (negative incentives)? * How far did the choices made reflect (perhaps imagined) experiences with regulating past new and similar technologies (e.g. privacy)? * And lastly, what role did the ideological preferences of the parties in government play in those choices? Based on evidence from the two case studies the paper argues that policy decisions on GenAI have rather less been based on rational evaluations and decisions between the competing goals and rather more on experiences with past regulatory decisions and by characteristics of the broader political system which create opportunity structures. The perception of enormous time pressure has considerably contributed to that. Proposed panel: P1 Regulating new technologies, chaired by: Elena Escalante Block & Trym Nohr Fjørtoft