The rapid shift to hybrid work environments—where employees alternate between remote and on-site work—has fundamentally altered organizational risk landscapes. Insider threats have grown more complex due to distributed access, blurred work–life boundaries, unmanaged personal devices, and weakened physical oversight. This paper synthesizes major insider-threat detection theories and adapts them to hybrid contexts, offering a conceptual model integrating behavioral, psychological, socio-technical, and AI-driven detection mechanisms. The study proposes a Hybrid Insider Threat Detection (HITD) framework combining Behavioral Theory, Zero-Trust Architecture, Routine Activity Theory, and Machine-Learning-Based Anomaly Detection. Through systematic literature analysis, the study evaluates how hybrid work reshapes antecedents, indicators, and mitigation strategies for malicious and unintentional insider threats. Practical implications, research gaps, and future avenues for designing intelligent insider-threat monitoring systems are discussed.