Prof. Siwei Lyu Warns Lawmakers About AI Scams and Data Misuse
Prof. Siwei Lyu, a SUNY Distinguished Professor at the University at Buffalo, testified before the New York State Assembly to warn lawmakers about the growing risks posed by AI-enabled scams.
Siwei Lyu explained that while traditional data mining already reveals sensitive details about people’s lives, using this data to train generative AI greatly increases the danger. Lyu emphasized that personal images and voice recordings collected without proper safeguards can be turned into highly realistic deepfakes that easily deceive consumers, creating serious threats to privacy, security, and trust.
Drawing on his expertise in digital media forensics and AI, Lyu urged lawmakers to adopt stronger protections for the public. He highlighted his work co-directing UB’s Center for Information Integrity and developing the widely used DeepFake-O-Meter, which helps detect manipulated digital content. Lyu recommended clear transparency rules for data collection, limits on how much data companies can gather, independent AI audits, reliable tools to detect AI-generated media, and greater investment in public education—especially for vulnerable groups. While praising New York’s leadership through the Empire AI initiative, he cautioned that unregulated data practices and generative AI misuse remain a serious and growing risk.
