Safeguarding Trust in the Age of Generative AI

Speaker:       Chee Wei Tan

Time:            10:30 am, Dec. 1st.

Location:      SIST 1A-200

Host:             Prof. Ziping Zhao

Abstract:

The rise of generative AI presents unprecedented challenges for verifying information. While AI hallucinations pose risks, the greater danger lies in human-augmented misinformation—where users may present AI outputs as breakthroughs without disclosing whether AI models discovered obscure references or simply fabricated results. This threatens the integrity of scientific communication and parallels threats in software supply chains, where malicious code can masquerade as trusted software repositories. This talk addresses these challenges through computational verification frameworks. We demonstrate how graph algorithms for large-scale optimization can analyze scientific publication networks to detect anomalous citation patterns and identify potentially unreliable claims. Simultaneously, we argue that trustworthy AI systems must integrate verification-focused AI copilots, rigorously testing AI model outputs to maintain integrity and ensure these tools build trust rather than erode it.

Bio:

Dr. Chee Wei Tan is with the College of Computing and Data Science, Nanyang Technological University, Singapore. He received his Ph.D. in Electrical Engineering from Princeton University. His research focuses on distributed optimization, Generative AI, networks, and edge computing. Dr. Tan is serving or has served as Symposium Co-Chair of 2025 IEEE Globecom Symposium on AI-Enabled Networks, Editor for IEEE Transactions on Signal and Information Processing over Networks, IEEE Transactions on Networking, IEEE Transactions on Communications, IEEE Transactions on Cognitive Communications and Networking, and IEEE Communications Society Distinguished Lecturer. He was selected twice for the U.S. National Academy of Engineering China-America Frontiers of Engineering Symposium.