Human centered artificial intelligence driven cybersecurity for business resilience using bibliometric and qualitative analysis

Cyber threats have become increasingly automated, adaptive, and difficult to manage through conventional rule-based security mechanisms. Artificial intelligence, machine learning, and data-mining techniques offer advanced capabilities for detecting suspicious activities, intrusions, malware, phishing, and anomalous network behavior. However, existing research remains predominantly model-centered and provides limited understanding of how AI-enabled cybersecurity systems are designed, explained, trusted, and used in practice. This study investigates AI-driven cyber-threat detection and chatbot-supported cybersecurity through a sequential mixed-method design. The first stage conducted a bibliometric analysis of publications indexed in Scopus and the Web of Science Core Collection from 1 January 2016 to 30 April 2026. The 2026 records represent a partial publication year. The analysis examined publication trends, contributing authors, countries, sources, citation patterns, and keyword clusters. The second stage involved semi-structured interviews with 10 developers and 10 users to explore the functionalities, implementation requirements, perceived benefits, and concerns associated with chatbot-supported cybersecurity. The bibliometric findings identified three major research areas: cybersecurity and threat-detection applications, artificial intelligence and machine-learning detection methods, and data-mining-based network-intrusion detection. The qualitative findings indicate that cybersecurity chatbots can provide rapid warnings, explain cyber risks, support phishing identification, offer first-level incident guidance, improve cybersecurity awareness, and escalate serious cases to human specialists. However, participants also identified concerns regarding data quality, privacy, explainability, trust, false positives, false negatives, adversarial manipulation, and overreliance on automated advice. The study proposes a human-centered framework integrating data foundations, detection intelligence, chatbot interaction, developer responsibility, user acceptance, governance, human oversight, and continuous learning. The findings position cybersecurity chatbots as supportive communication and decision-assistance tools rather than replacements for validated detection technologies or cybersecurity professionals.