PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
April 3, 2026Open Access

Characterizing AI-Enabled Social Engineering Threats to U.S. National Security: A Mixed-Methods Analysis for Developing Defensive Policy and Technical Countermeasures

View Full Paper
Ask AI
Bookmark
Share

Authors

LPLaszlo Pokorny

Discussion

Loading...

Member takes

Overview

Mixed-methods analysis identifies AI social engineering threats and gaps in defensive policies.

Key Points

  • This research aims to characterize AI-enabled social engineering threats and evaluate defensive measures and policy preparedness.
  • Conducted a mixed-methods analysis combining quantitative and qualitative approaches.
  • Analyzed large-scale phishing email datasets using machine learning classification to identify AI-generated content.
  • Conducted thematic content analysis on cybersecurity policy documents like NIST Cybersecurity Framework and CISA guidelines.
  • TF-IDF Logistic Regression model achieved 92% accuracy and an AUC-ROC of 0.9869.
  • Structural features were more predictive of AI-generated phishing than linguistic features.
  • Identified five critical policy gaps in current frameworks regarding AI threats.

Cite This Study

Laszlo Pokorny (2026) studied this question.

synapsesocial.com/papers/69cf5f645a333a821460e7bdhttps://doi.org/10.5281/zenodo.19354428
View Full Paper
Ask AI
Bookmark
Share