Prof. Callahan and Michael Ivanicki ’25 Publish on Phishing Detection Systems

Brian Callahan, Ph.D., specialist professor in the Department of Computer Science and Software Engineering, and Michael Ivanicki '25, graduate computer science student, recently collaborated to present at the 14th International Symposium on Digital Forensics and Security (ISDFS 2026) in Boston, MA. Their work was conducted through the Monmouth University Cybersecurity Research Center and the corresponding paper, "An Explainable AI Study of Phishing Detection Model Degradation Across Real-World Events," was published by IEEE Xplore.

The authors examined how phishing detection systems can lose effectiveness as cybercriminal tactics evolve over time. To conduct the study, Ivanicki and Callahan analyzed more than 300,000 phishing and legitimate websites collected between 2018 and 2020. Next, they trained machine learning models on data from individual years, then tested how well those systems performed on phishing attacks from later years.

The researchers found that phishing detection models degraded significantly over time as attackers changed their methods. They concluded that shifts in attacker behavior during major events such as the rollout of 5G infrastructure and the COVID-19 pandemic contributed to changes in phishing techniques that reduced the reliability of older detection systems.

The authors said the findings highlight the need for cybersecurity systems that can adapt to changing conditions rather than relying on static training data. They also said future research will examine phishing trends after 2020 to better understand how attacks continue to evolve in response to global events and technological change.

"Working on this research provided a valuable opportunity to apply concepts that I learned about in class to a real-world project," said Ivanicki. "Collaborating closely with Dr. Callahan not only strengthened the technical direction of the project, but it also granted me the opportunity to present our work at an international cybersecurity conference. Through this experience I gained a deeper understanding of how AI can be utilized to benefit cybersecurity, and how real-world events heavily influence the cyber threat landscape."

The ISDFS conference offers a platform for researchers and professionals from academia, industry, and government to exchange ideas and recent advancements in digital forensics, cybersecurity, and computer science more broadly. The 14th ISDFS conference featured a diverse program including special sessions, workshops, tutorials, keynote speeches, panel discussions, posters, and oral presentations.