As Large Language Models (LLMs) generate increasingly sophisticated text-image pairs, the risk of LLMs being used for phishing is growing, posing significant challenges for Internet security. We conduct a detailed, IRB-approved study of pairs (uc,pc) where uc is a user property (e.g., demographic or behavioral) and pc is a property of the post content (e.g., topic, emotion, and recency). Unlike past research that focuses either on user properties alone or post content alone, we conduct a comprehensive statistical analysis of more than 1,400 hypotheses involving such (uc,pc) pairs and identify the conditions on user-content pairs that have a significantly higher click probability compared with pairs that do not satisfy the condition. The results include highly nuanced findings connecting susceptibility to phishing with the combination of user properties and post content properties. They also cast some light on contradictory findings from prior work on understanding human susceptibility to phishing.

Fake audios, videos, and images are now proliferating widely. We developed GODDS, the Global Online Deepfake Detection system, for a specific user community, namely journalists. GODDS leverages an ensemble of deepfake detectors, along with a human in the loop, to provide a deepfake report on each submitted video/image/audio or VIA artifact submitted to the system. To date, VIA artifacts submitted by over 50 journalists from outlets such as the New York Times, Wall Street Journal, CNN, Agence France Press, and others have been run through GODDS. Unlike other deepfake detection systems, GODDS doesn't just focus on the submitted artifact but automatically derives context about the subject of the VIA artifact. Because context is not always available on all subjects, GODDS focuses on alleged deepfakes of high profile individuals, organizations, and events, where there is likely to be considerable contextual information.

We consider behavior of agents in a long-term multiagent coopetitive setting in which agents vary their cooperative and competitive stances over time. Using the game ofDiplomacyas a testbed, we study how successful agents vary their coopetitive behavior, developing a new “style of play” (SoP) characterization of player behavior. We assess five novel SoP hypotheses about successful behavior. We propose two algorithms to automatically compute an agent’s SoP vector and describe the important factors in this computation. As an agent’s SoP depends on the game state and its perception of threat, we develop a novel “means, motive, and opportunity” (MMO) model of threat and show that we can predict threats effectively using this model. We provide novel insights into how agents should behave to more successfully achieve their goals in long-term coopetitive settings.