Post

Describing attacks with crime script analysis

Crime script analysis (CSA) is a narrative-driven methodology from criminology that decomposes attacks into discrete sequential steps and situational requirements, offering a complement to MITRE ATT&CK for communicating threats to non-technical audiences. Using business email compromise (BEC) as a case study, the article demonstrates how AI automation is shifting BEC from high-value fraud against few targets to lower-value fraud against many targets by automating reconnaissance (target identification, personnel discovery, organizational context) and social engineering personalization. The article identifies five specific intervention points where defenders can disrupt the attack: seeding fictitious canary organizations to detect automated reconnaissance, monitoring LLM interactions for suspicious reconnaissance patterns, rate-limiting anomalous email flows at the sending stage, blocking discovered malicious sources, and victim awareness training. CSA provides a practical framework for defenders to identify where attackers are applying AI tools to scale attacks and where strategic disruptions can be deployed to break the attacker’s workflow.

Read original article

This post is licensed under CC BY 4.0 by the author.