The Promise and Perils of AI-Mediated Political Communication

Smart replies, writing enhancements, and virtual assistants powered by artificial intelligence language technologies have been increasingly integrated into consumer products and everyday experiences.  This research explores the potential and risks of AI-mediated communication (AI-MC) technologies such as GPT-4 in the political space through a series of experiments that explore both the potential uses and misuses.  The first part of the research evaluates whether human-AI collaboration can increase legislator responsiveness by studying citizen responses to AI-generated tweets and email correspondences. Here the research highlights the importance of disclosure, transparency, and human-in-the-loop accountability for AI-mediated political communication. The second part of the research addresses the plausibility of misuse by actors seeking to influence the democratic process. It reports the results from a field experiment on legislators, shows the challenge these technologies present to democratic representation, and finally suggests potential techniques that elected officials might employ to guard against AI-sourced astroturfing.

Advance permission required to attend in person

Speaker: Sarah E. Kreps, Cornell University