AI Policy Development: Leveraging Social Science to Ensure Responsible Generative AI

How should frontier AI companies decide what their models may enable, who should have access, and when the state or public should have a say? In this policy practicum, students work as product-policy advisors to a frontier AI company, drawing on social science research, political theory, and historical cases to develop feasible recommendations for real AI policy problems. The course treats decisions about model behavior, access, deployment, privacy, safety, and enforcement as a form of governance that allocates power among companies, citizens, and the state. Historical cases including the printing press, nuclear technology, railroads, radio, pharmaceutical regulation, algorithmic trading, and the Crypto Wars illuminate recurring dilemmas that students apply to contemporary AI products. Students produce policy memos and participate in a multi-stakeholder simulation based on a current AI product-policy question. This course is a Capstone course for the Political Science Major. Enrollment in the class is prioritized for Political Science Seniors. Completion of POLISCI 1: The Science of Politics is required prior to enrollment.