CAREER: Leveraging Scambaiting Community Knowledge to Defend Against Interactive Social Engineering Attacks

Online scams increasingly rely on sustained, real-time conversations in which attackers manipulate victims through multiple communication channels. These interactive scams cause billions of dollars in losses each year. However, effective detection systems are difficult to build because little data exists about how these conversations unfold. The interactions typically occur privately between attackers and victims, making them difficult for researchers to analyze. One of the project’s novelties is the systematic study of knowledge generated by the global scambaiting community, individuals who deliberately engage scammers in conversations to waste their time. Another novelty is combining insights from this community with controlled research methods to advance understanding of how scammers persuade victims during interactive attacks. The project's broader significance and importance are advancing society’s ability to detect and disrupt online scams while improving public awareness of how interactive social engineering attacks operate. The project also engages the wider community through educational workshops that raise awareness of scam tactics and strengthen resilience against online fraud. This project develops a research framework for measuring and defending against interactive social engineering attacks on the internet. The research analyzes practices of the scambaiting ecosystem to study tools and techniques used by scambaiters during real-time engagements to keep conversations with scammers active. Building on these insights, the project develops controlled systems that engage scam perpetrators and collect datasets of conversational interactions while incorporating ethical safeguards. These datasets support detection methods that leverage conversational patterns, behavioral signals from computer activity, and voice characteristics to identify scam attempts. In addition, the research examines how detection signals should be presented through user interfaces to minimize false alarms while maintaining usability. The expected outcomes include new measurement methodologies, improved scam detection techniques, publicly disseminated research findings, and educational resources that strengthen community resilience against online fraud. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria. NSF Award ID: 2544625 | Program: 01003031DB NSF RESEARCH & RELATED ACTIVIT,01002627DB NSF RESEARCH & RELATED ACTIVIT,01002930DB NSF RESEARCH & RELATED ACTIVIT | Principal Investigator: Krishna Phani Vadrevu | Institution: Louisiana State University, BATON ROUGE, LA | Award Amount: $388,063 View on NSF Award Search: https://www.nsf.gov/awardsearch/show-award/?AWD_ID=2544625 View on Research.gov: https://www.research.gov/awardapi-service/v1/awards/2544625.html