Understanding Social Engineering Attacks with Natural Language Processing
UC Irvine professor Ian Harris demonstrates using Natural Language Processing (NLP) and syntactic parsing to automatically detect social engineering intent.
Summary
The Linguistic Structure of Social Engineering
Unlike technical software exploits that rely on binary syntax errors, social engineering attacks rely on structured linguistic pretexts, urgency cues, and emotional coercion.
Parsing Speech Acts and Semantic Intention with NLP
Ian Harris details using semantic role labeling and dependency trees to identify conversational requests, authority assertions, and commitment elicitation.
Building Automated Text Analysis Classifiers
Harris demonstrates training machine learning models on transcripts of SECTF phone calls and malicious email corpuses to score social engineering probability.
Future Applications in Automated Threat Defense
The author discusses deploying real-time NLP filters on enterprise chat clients, email gateways, and customer support ticketing queues.
Related talks
Main Stage·26 min
Eugene Lim · Glenice Tan · Tan Kee Hock
Singapore government researchers construct an automated spear-phishing pipeline powered by OpenAI's GPT-3 API, comparing AI-generated phishing to human campaigns.
Blacks in Cybersecurity Village·33 min
Levone Campbell
Explores the dual role of artificial intelligence in scaling hyper-targeted phishing campaigns while powering real-time behavioral defenses.
AI Village·49 min
Lee · Saxe
Presents a cutting-edge deep neural network architecture designed to detect subtle social engineering patterns in emails.