School of CSE Seminar Series: Kai Shu

Speaker: Kai Shu, assistant professor at Emory University
Date and Time: April 9, 11:00 a.m. - 12:00 p.m.
Location: Coda 114
Host: TBA

Title: Combating Misinformation in the Age of Generative AI: Challenges, Risks and Opportunities

Abstract: In recent years, misinformation and disinformation have become a defining challenge for civic life and the public information ecosystem. The wide dissemination of misinformation can have detrimental societal effects, dividing people, polarizing groups, confusing readers, and being weaponized by nation-states to exert geopolitical influence. Over the past decade, researchers have developed methods for detecting misinformation on social media by leveraging social context, network signals, and cross-domain learning, yet fundamental challenges around data scarcity, explainability, and reliability remain.

The rise of generative AI, particularly Large Language Models (LLMs), has the potential to reshape this fight. However, LLMs are a double-edged sword. While they offer promising opportunities due to their extensive knowledge and reasoning capabilities, enabling zero-shot fact-checking, explainable claim verification, and authorship attribution, they also pose significant risks to public trust and online safety, as they can be exploited to generate highly deceptive misinformation at scale. In this talk, we explore lessons learned in harnessing LLMs for identifying misinformation, discuss the challenges of detecting LLM-generated misinformation, and emphasize the need for interdisciplinary collaboration across technologists, social scientists, journalists, and policymakers to address the complexities of combating misinformation in the age of generative AI.

Bio: Dr. Kai Shu is an Assistant Professor of Computer Science at Emory University, specializing in responsible AI, machine learning, data mining, and natural language processing, with applications in social computing, healthcare, and cybersecurity. He earned his Ph.D. in Computer Science from Arizona State University under Dr. Huan Liu. His research has been recognized with numerous awards, including the NSF CAREER Award, Microsoft Accelerate Foundation Models Research Award, DARPA AI Forward Scholarship, AAAI New Faculty Highlights, Cisco Faculty Research Awards, and Dean’s Dissertation Award. He is the author of Detecting Fake News on Social Media (Morgan & Claypool, 2019) and has developed FakeNewsNet, a widely-adopted benchmark repository for fake news research on social media. His work has appeared in leading venues such as ICLR, NeurIPS, KDD, WWW, EMNLP, and AAAI, and is funded by DARPA, IARPA, the Department of Homeland Security, the Department of Energy, and the National Science Foundation, as well as industry partners including Google, Cisco, and Microsoft. He collaborates actively with social scientists, health informaticists, and members of the intelligence community to bridge computational approaches with real-world policy impact. For more information, visit https://www.cs.emory.edu/~kshu5/.