AI in the German Bundestag: On the Relationship between Evidence-Based Policymaking and Artificial Intelligence

Artificial Intelligence (AI) as a general-purpose technology is widely considered to be one of the next major drivers of digital transformation. As a technology capable of processing vast amounts of data and generating information far beyond human capabilities, it is an excellent example to discuss evidence-based policymaking and its current evolution. AI is seen as transforming all aspects of our society, and while we've already seen some well-established and not-so-well-established use cases in recent years, the political debate around the technology and its promises and challenges only became virulent in 2017 and 2018, when several countries released their own national AI strategies. There is already a steadily growing, substantial debate about AI regulation, mostly focused on national governments or supranational governance options. But also, about necessary funding options and transfer options from research to applications. What is missing in most cases is a parliamentary perspective on these developments. Although there are some initial theoretical and abstract reflections on how AI alters democratic political systems in general, and what positive and negative scenarios are conceivable in terms of the throughput level, we know little about how AI is already affecting parliamentarians and their policymaking. If AI is understood as a technology that uses data to generate information, which in turn could possibly lead to knowledge, there are several conceivable paths of affection. This paper focuses on the implications of the growing number of AI applications and their possible effects on the parliamentary work of MPs by asking: How do parliamentarians use the possibilities of AI and how does this show? Designed as a single case study, the paper examines how German MPs use AI and its possibilities in terms of evidence-based policymaking, in favor of new ways to extract information from unimaginable amounts of data, or even to argue against these possibilities by rejecting the technology. The paper uses two main sources of research data: First, semi-structured expert interviews with former members of the AI Inquiry Committee are analyzed using qualitative content analysis. The selection of these MPs is based on the assumption that the members of the commission can be considered as a kind of forerunners. Second, parliamentary debates of the German Bundestag will be used to identify patterns of how MPs use AI to argue for their own policy positions or against those of political opponents. The research period begins with the 19th legislative period, as it can be argued that this period (2017 - 2021) marks a turning point for the topic of AI within the German parliament, with the installation of the first parliamentary commission of inquiry on the topic, but also in a broader sense, with references to the topic in the coalition agreement of the new government, a separate national AI strategy, as well as an ethics committee for AI. In doing so, the paper sheds light on how AI, as a prominent digital technology, is being used by parliamentarians to promote a next level of evidence-based policymaking or is even being rejected.