Developing an Anti -Phishing Large Language Model: A Focus Group Study on Human, Technological, and Legal Factors
George A. Thomopoulos, Panagiota Kiortsi, Damianos Dumi Sigalas, Dimitrios Kosmopoulos, Ioannis Inglezakis, Christos A. Fidas · 2024
The phishing problem poses a significant threat in modern information systems, putting both individuals and businesses at risk of financial and professional harm. Owing to social media's rapid development and widespread appeal, deception of this sort is hurting millions of people and growing more dangerous. The current work, as part of the AILA (Artificial Intelligence-driven Framework and Legal Advice Tools for Phishing Prevention and Mitigation in Information Systems) project, aims to specify and validate an AI-driven multifactor (human, technology and legal) anti-phishing data model, with the implementation of focus group studies. The findings assist to provide human, technology, and legislative user model endpoints that will be identified and discussed for explicit and implicit user modeling, which will guide the development of the corresponding AI-driven user modeling and profiling mechanisms. To this end a Large Language System is planned to be employed.