The insurance sector has witnessed a paradigm shift with Artificial intelligence (AI), revolutionizing risk management, customer engagement, and operational workflows. This study synthesizes global research trends, innovations and unresolved challenges in AI-driven insurance risk management through a dual-method approach: bibliometric analysis of 329 Scopus and Web of Science publications (1988-2025) and systematic literature review (SLR) of 60 rigorously filtered studies. Key applications span fraud detection, claims automation, personalized underwriting and customer service optimization. However, critical gaps persist, including regional imbalances in research focus, limited exploration of niche insurance domains, and ethical concerns like algorithmic bias. The study proposed policy interventions to mitigate data scarcity, regulatory misalignment and skill shortages while advocating interdisciplinary research integrating AI with blockchain and IoT for enhanced risk prediction. By mapping influential authors, institutions and thematic clusters, this work equips insurers, transformative potential responsibly, fostering innovation without compromising equity or transparency in risk mitigation strategies.