AI-Enhanced Phishing Attacks: Understanding the Threat Landscape and Defense Strategies

The increasing accessibility of generative AI tools has fundamentally reshaped the landscape of phishing threats. These tools enable attackers to produce context-aware personalized phishing content that can bypass both traditional defense mechanisms and human cognitive filters. This paper aims to help researchers understand the evolving threat. We propose a novel survey of the state of AI-enhanced phishing attacks and compare them with traditional tactics in key dimensions, including threat surface expansion and changes in attack life-cycle. Recent defense strategies are further examined as technical, non-technical, and hybrid approaches. In addition, we conducted a case study to assess how highly educated individuals respond to AI-generated text-based phishing attacks. The results show that 61. 1% of the participants identified phishing indicators, while 38. 9% failed to detect the phishing attempt. Future research should explore real-time defense mechanisms that leverage user intent modeling, sentiment analysis, and cross-platform AI integration.