Suggested citation: Chen, Nuole. 2025. MIT GOV/LAB Generative AI & Democracy, SERC Reading Group. Massachusetts Institute of Technology, Department of Political Science.

Overview

We live in an era where technology has changed the ways that we interact in a democracy, and generative AI accelerates that change.  Kenya’s protests over the 2024 Finance Bill gives powerful examples of how generative AI has become part of democratic action.  During the protests, custom GPT tools like the “Finance Bill GPT” and “Corrupt Politicians GPT” were developed.  The Finance Bill GPT gives people easy access to a difficult-to-digest government bill, and the Corrupt Politicians GPT gives people access to corruption scandals and cases of politicians.  In India, generative AI is used to enable access to public services and to process high volumes of public grievance redressals in multiple languages.  

This impact paper discusses the many ways that generative AI can bolster and improve deliberative platforms for strengthening democracy and researchers have found that generative AI can increase civility online.  With generative AI, information to promote accountability, transparency, and awareness of issues can become much more accessible and available.  Governments also have better tools for responding to citizens and providing public services.  And citizens have improved tools to have discourse, express themselves, and make sure their views are represented.

At the same time, generative AI can easily pose threats to democracy.  Just as it is a source of more accessible information, it is also a source for misinformation.  During this presidential campaign season in the US, fake AI-generated audio of President Biden discouraged New Hampshire citizens from voting.  In fact, AI-powered misinformation campaigns can easily flood social media and the information ecosystem, and it is only becoming cheaper and faster.  People are suspicious of generative AI, but as it is used in more government transactions or used in deliberation platforms, growing distrust of government or discourse can also emerge. 

This syllabus provides an opportunity to delve deeper into how generative AI can impact democracy, in ways that can strengthen democracy and that are socially and ethically responsible.  The readings cover several topics related to generative AI and democracy to  foster discussion around a healthy democracy and how generative AI affects them.  This resource will also be updated as new research about generative AI and democracy come out.  For suggestions of readings to add to this syllabus, please reach out to govlab@mit.edu.

Photo by Google DeepMind on Unsplash.