A Comparative Study of Trust in AI-Driven Security Alert Systems and Human Cybersecurity Experts in IT Environments

Abstract

The proliferation of automated cyber security tools including antivirus software, browser security warnings, email phishing alert systems, and workplace security platforms has fundamentally altered how individuals encounter and respond to digital threats. Concurrently, human security experts continue to provide context-sensitive advisory services, raising critical questions about how end users comparatively trust these two advisory channels. Despite growing organizational reliance on automated security systems, limited empirical evidence exists regarding how non-expert and semi-technical users perceive, compare, and act upon automated security alerts versus advice from human security experts. Understanding this trust dynamic is essential for designing more effective cyber security systems and education programmes. This study employed a cross-sectional quantitative survey design. Data were collected from 31 valid respondents using a structured 5-point Likert-scale questionnaire organized into five sections covering demographics, automated trust, human expert trust, comparative preferences, and behavioral intentions. Participants were primarily undergraduate students aged 18–24 with varying technical backgrounds, recruited via Google Forms during April 2026.Findings reveal moderate trust in both automated alerts (Composite M = 3.50, SD = 0.73) and human experts (Composite M = 3.39, SD = 0.81). When sources conflicted, 77.4% of respondents preferred to seek additional information rather than defer exclusively to either advisory channel. In high-risk scenarios, 51.6% favoured relying on both sources equally, 32.3% preferred human experts, and only 16.1% preferred automated systems alone. Users occupy a deliberative middle ground, valuing both the speed and consistency of automated alerts and the contextual judgment of human experts. The findings carry implications for cybersecurity interface design, security education, and institutional advisory policy.