Abstract
Artificial intelligence (AI) is transforming cybersecurity by strengthening defensive capabilities while simultaneously enabling more sophisticated phishing attacks. AI generated phishing emails increasingly evade traditional detection methods through improved grammar, contextual relevance and realistic communication styles creating significant challenges for Security Operations Centers (SOCs). This study investigates the effectiveness of current email security systems in detecting AI generated phishing emails compared with traditional phishing emails. A mixed method experimental approach was employed using controlled Gmail and Outlook environments to evaluate the detection and classification of multiple phishing scenarios. The results demonstrate that several AI generated phishing emails successfully bypassed initial security filters and reached users' inboxes before detection or remained undetected entirely. These findings highlight limitations in conventional rule based filtering approaches and emphasize the need for behavior based detection, AI assisted threat analysis, enhanced employee awareness training and layered security strategies within SOC environments. This research contributes practical insights into the growing threat of AI enabled phishing attacks and provides recommendations for improving organizational detection and response capabilities.
Included in
Information Security Commons, Management Information Systems Commons, Technology and Innovation Commons