Human-AI Collaboration in Cybersecurity Decision-making: A Systematic Review of Literature

Abstract

: Artificial intelligence has become an important tool in cybersecurity decision-making, but there are arguments about balancing it with human cognition. Human cognition can significantly influence cybersecurity outcomes, making it essential to examine human–AI interaction within cybersecurity environments.  Thus, this study examined human-AI collaboration in cybersecurity decision-making using systematic review of literature. Using the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA), a total of seventeen (17) articles were selected from credible databases using some set of inclusion and exclusion criteria. The study findings showed that human-AI collaboration in cybersecurity decision-making is a complementary and symbiotic approach wherein AI enhances human judgement through structured frameworks and architectures. Results showed that AI performs real-time monitoring and analysis while humans handle complex or high-risk decision, supported by dynamic role-based models that allow flex collaboration between human and AI. Findings indicate that collaborative human-AI cybersecurity decision-making is supported by a combination of technical and organizational techniques, which enhance transparency, trust, accuracy, and alignment between human judgement and AI outputs. Findings showed that as a result of its efficiency, scalability, and adaptability, human-AI collaboration in cybersecurity decision-making is more effective than human-only or AI-only approaches. The study concludes that despite the challenges of using human-AI collaboration in cybersecurity decision-making, it is more effective than using solo approach – whether human or AI only.