MITIGATING AI-POWERED PHISHING ATTACKS THROUGH ADAPTIVE EMPLOYEE TRAINING AND BEHAVIOURAL ANALYTICS

The increasing use of artificial intelligence (AI) in cybercrime has significantly enhanced the sophistication of phishing attacks, enabling adversaries to generate highly personalized and convincing social engineering campaigns through machine learning, natural language generation (NLG), and deepfake technologies. This study investigates the effectiveness of combining behavioral analytics with adaptive staff training as a human-centered approach to mitigating AI-powered phishing attacks. The study adopts a case-study and evidence-based analytical approach to examine behavioral indicators such as phishing click rates, credential-submission behavior, reporting rates, and repeated susceptibility to simulated phishing attacks. Findings from evaluated adaptive cybersecurity training interventions demonstrate substantial improvements in employee phishing resilience. AI-driven adaptive training has been associated with a reduction in simulated phishing click rates from 8.8% to 2.1%, representing a 75% relative reduction, compared with minimal improvement in a control group. Related behavioral interventions have also demonstrated reductions in phishing click rates from 27.6% to 14.2% and credential-submission rates from 11.3% to 4.9%, while phishing reporting rates increased from 18.9% to 36.5%. These findings demonstrate that continuous behavioral monitoring can identify high-risk users and enable training interventions to be adapted according to individual susceptibility and risk profiles. The study concludes that integrating behavioral analytics with adaptive cybersecurity training provides a more proactive and measurable approach to human risk management than conventional one-size-fits-all awareness programs. Continuous behavioral assessment, personalized training, and real-time feedback can therefore strengthen organizational resilience against increasingly sophisticated AI-enabled phishing and social engineering attacks.