Between You and the Facts

Join the Tech Impact and Policy Center on October 20th from 12PM–1PM Pacific for a seminar with Matt Deverna.

This seminar will be held in-person only. We will not have a livestream or webinar option for this talk.

Stanford affiliates are invited to join us at 11:40 AM for lunch, prior to the seminar. The Fall Seminar Series continues through December; see our Fall Seminar Series page for speakers and topics. Sign up for our newsletter for announcements.

About the Seminar:

People increasingly turn to AI systems to make sense of the news and to answer questions about their health. What happens when the answers we get about politics and our own health come from AI? In this talk, I will present two projects on AI and political fact-checking. The first is an experiment testing how AI-generated fact-checks of political news headlines affect people's ability to tell true headlines from false ones. The second is an audit of how well 15 large language models fact-check 6,000 political claims rated by PolitiFact. I will also share ongoing work examining how six AI systems (Google's AI Overviews, Gemini, ChatGPT, Claude, Grok, and Perplexity) respond to general health questions. Together, these studies shed light on when AI helps people separate fact from fiction, where it falls short, and what this means for the growing number of people who rely on it.

About the Speaker:

Matt DeVerna is a computational social scientist studying the societal impacts of artificial intelligence, with a focus on how people find information online to form beliefs and make decisions. His research spans human-centered and system-level analysis, pairing large-scale audits of AI systems, such as language models and search engines, with experiments and observational analysis of social media. This work centers on high-stakes domains like news, politics, and public health, where these systems can strengthen or erode both public understanding and public institutions. It aims to help researchers, platforms, and policymakers ensure AI makes our digital spaces more trustworthy and resilient. Matt is a postdoctoral scholar at Stanford University’s Tech Impact and Policy Center. He holds a Ph.D. in Informatics from Indiana University and an M.A. in Psychology from New York University.