(Re)distributive Politics after the AI Revolution

forthcoming

Recent advancement in Artificial Intelligence (AI), such as the release of ChatGPT, has intensified concerns about the future of work. Studies suggest that ChatGPT could already fully replace up to 33% of jobs, with high-skilled workers facing the greatest risk. Scholars warn that AI automation may threaten the central role of work as the primary means of income distribution, potentially leading to an era of widespread poverty and intensified inequality. In response, many countries are seeing growing political momentum to reform welfare in anticipation of rapid changes in the world of work. Yet, political science has yet to examine how welfare states will respond and adapt to the labor market challenges posed by AI.   This project is the first comprehensive analysis of welfare state politics in the age of AI, examining how institutions, and the supply and demand of welfare politics shape responses addressing AI's impact on the labor market. Employing a multimethod approach that integrates natural language processing and multilevel models, it aims to address the central question: How does AI reshape welfare politics across OECD countries?   The project will examine in real-time how welfare policies and politics are evolving in response to the rapid changes brought by AI. The project will: (1) shed light on different welfare systems will be affected by AI, (2) map the political landscape of party positions and framings of AI, (3) examine public support for different welfare reforms, (4) develop a new understanding of how welfare systems will adapt to the challenges posed by AI. By exploring how existing systems shape the adoption of AI, the project moves away from techno-deterministic perspectives to focusing on the active role of policies in shaping the future. The project breaks new academic ground by developing a new area of study, and offers actionable recommendations to policymakers on viable paths for welfare reforms to address the challenges posed by AI.