Context-Aware Spear Phishing: Generative AI-Enabled Attacks Against Individuals via Public Social Media Data

2026-05-11

AI Summary

This work proposes a context-aware spear-phishing framework leveraging generative AI and publicly available social media data, which for the first time systematically integrates multimodal signal extraction, user communication style modeling, and seven distinct attack strategies to automatically generate highly personalized and scalable phishing emails. The synthesized messages significantly outperform real-world samples in contextual relevance, personalization, and persuasiveness, with user studies indicating substantially lower perceived suspicion. The study further evaluates multiple state-of-the-art defense mechanisms, exposing critical limitations of current content moderation systems against contextualized, adaptive attacks, and introduces novel defensive paradigms such as prompt-level adaptive defenses and chain-of-thought auditing.

Abstract

We demonstrate how publicly available social-media data and generative AI (GenAI) can be misused to automate and scale highly personalized, context-aware spear-phishing campaigns. With minimal attacker effort, a small amount of public activity per target is sufficient for GenAI models to extract interests and contextual cues, producing persuasive messages that mirror a target's style while bypassing generic content-moderation safeguards. We introduce a modular framework that combines multimodal signal extraction, communication-style profiling, and attack-type instantiation across seven strategies (baiting, scareware, honey trap, tailgating, impersonation, quid pro quo, and personalized emotional exploitation). We conduct a large-scale, multi-model evaluation covering thousands of generated emails and eight security-relevant criteria, benchmarking against a corpus of real-world phishing messages. The GenAI-produced emails exhibit markedly higher personalization, contextual grounding, and persuasive leverage. Importantly, a complementary user study corroborates these results, revealing that LLM-generated attacks consistently outperform APWG eCrimeX emails across eight dimensions while eliciting lower suspicion among human recipients. Finally, we measure and analyze the behavior of existing proactive, prompt-level defense mechanisms, which incorporate adaptive mechanisms, as well as two complementary defense approaches-policy-augmented SOTA safeguard models and system-instruction chain-of-thought moderation. We document how these defenses respond to contextualized and adaptive attack prompts, underscoring the need for platform-level safeguards that explicitly account for contextualized abuse at scale.

Problem

• spear phishing
• generative AI
• context-aware attacks
• social media data
• personalized cyberattacks

Innovation

• context-aware phishing
• generative AI
• personalized cyberattacks
• multimodal signal extraction
• adaptive defense mechanisms