Paper short abstract:

We offer an empirically grounded account of the current functions of AI discourse and the promise of AI for political systems and interpret these findings in the light of the crises of contemporary capitalism. We underpin our thesis with empirical data on public opinion on AI in Germany.

Paper long abstract:

It is a commonplace theme to lament the potential long-term dangers of Aritificial Intelligence (AI) for democratic states. Less often the potential of AI to deepen democracy is emphasized. But even more rarely the actual, contemporary effects of AI technology and AI discourse on democratic systems are analyzed. Unlike speculations about future courses of development, we can do empirical social scientific research on the latter question. In our contribution we offer an empirically grounded account of the current functions of AI discourse and the promise of AI for political systems and interpret these findings in the light of the crises of contemporary capitalism. We underpin our thesis with empirical data on public opinion on AI in Germany.

Against this backdrop, we interpret content-analytic findings from two distinct data sources. First, a set of articles (N = 2,073) was sampled from 12 national print, online and broadcast media that address digitization and AI in pre and post-election coverage on the 2021 German federal election. A quantitative content analysis of these provides insights into the political role assigned to the topics of digitization and AI, with particular focus on risk-benefit as well as problem-solution assessments, articulated needs for political action and arguments justifying their use. Second, a continuous collection of the coverage about AI is sampled since January 2021 from 34 German online and print media.

We thereby hope to show the ideological and legitimatroy function of AI for contemporary (german) democracy.