This article argues that the rise of the subscription economy is prompting a paradigm shift in the insurance industry, as consumption shifts from one-time ownership to continuous, personalized experience. While traditional insurance fixes coverage and terms at the point of contract and maintains limited contact with policyholders outside of claims, subscription-based insurance—or "Insurance-as-a-Service"—embeds insurance within a recurring payment relationship characterized by flexible digital enrollment, real-time coverage adjustment, and data-driven personalization. The article classifies subscription insurance into curation-type services, where consumers or insurers configure coverage bundles, and premium-benefit-type services, which combine coverage with embedded partnerships or value-added risk-management offerings, illustrating each with international cases such as Liberty Mutual, Lemonade, Asurion, igloo, Vitality, and Oscar Health. It then identifies the domestic regulatory barriers to adoption in Korea, spanning disclosure obligations, dark-pattern regulation, special-benefit prohibitions, underwriting design, and data governance, and proposes a phased roadmap—beginning with low-risk embedded models, advancing through regulatory-sandbox piloting of curation services, and culminating in institutionally supported value-added offerings. It concludes that subscription-based insurance can help insurers overcome stagnant growth, particularly in life insurance, by lowering entry barriers and sustaining lifecycle-long customer engagement.

This article argues that the rapid diffusion of generative and agentic AI is transforming AI from a business tool into core industrial infrastructure, creating new and more complex risks for insurers. AI-related losses are no longer limited to inaccurate predictions or hallucinated outputs; with agentic AI, errors can directly trigger wrong transactions, system changes, business interruption, liability claims, cyber incidents, regulatory violations, reputational damage, and even physical losses. The article classifies AI risks into technology and operational failure, cybersecurity and privacy, legal and ethical liability, reputation and organizational capability, and sustainability risks, then explains how these become insurance losses through triggers such as internal-control failure, direct AI execution, system contagion, external manipulation, rights infringement, and physical operation. It concludes that AI risk is not inherently uninsurable, but insurers must clarify liability across the AI value chain, strengthen underwriting around governance and control points, manage accumulation risk, refine AI exclusions, expand existing insurance coverage, and gradually develop stand-alone AI insurance as data and scenario-based pricing methods mature.

This article examines the institutional and strategic issues surrounding the adoption of artificial intelligence in the insurance industry, with a particular focus on the Korean insurance market. It argues that although AI has significant potential to enhance underwriting, claims handling, customer service, fraud detection, and asset management, its practical use remains constrained by regulatory ambiguity, fragmented data infrastructure, weak governance systems, and concerns over algorithmic accountability, privacy, fairness, and consumer protection. The discussion identifies major risk categories associated with AI adoption, including errors in decision-making, discrimination and bias, personal-information leakage, cybersecurity threats, and uncertainty over responsibility when automated systems cause harm. In response, the article proposes a policy framework that clarifies legal responsibility, strengthens data-sharing and verification systems, promotes model transparency, and establishes governance standards suitable for insurance-specific AI applications. It further suggests that insurers should adopt AI gradually across the value chain by prioritizing high-impact and low-risk use cases, building internal data and compliance capabilities, and coordinating technological innovation with regulatory readiness. Overall, the article concludes that sustainable AI transformation in insurance requires not only technological investment but also institutional reform, risk-control mechanisms, and strategic implementation aligned with consumer trust and market stability.

This article examines the rapid diffusion of artificial intelligence in the insurance industry and argues that AI is becoming a core operational technology across the entire insurance value chain, from product development and underwriting to sales, claims handling, and asset management. Beginning with the broader expansion of the global AI market and the transformation of financial services, the article explains that insurance companies can use AI to process large volumes of structured and unstructured data, improve risk assessment, personalize products, automate customer interaction, detect fraud, accelerate claims payments, and enhance investment decision-making. Through examples such as Munich Re’s underwriting platform, AXA’s AI-supported customer service, Ping An’s automated claims system, and domestic insurers’ AI-based asset-management tools, the article shows that AI can increase efficiency, reduce costs, and improve customer experience. However, it also emphasizes that Korean insurers still use AI more cautiously and fragmentarily than global leaders, mainly because of data restrictions, regulatory uncertainty, legacy systems, and risk concerns. The article concludes that insurers should move beyond isolated adoption and develop company-wide AI strategies that integrate data infrastructure, governance, risk management, and service innovation, while regulators should provide flexible and adaptive institutional support for responsible AI use in insurance.

This paper examines how climate change and the increasing frequency and severity of extreme weather events are transforming the insurance industry by amplifying disaster losses, changing the spatial and temporal pattern of risk, and widening the protection gap between total economic losses and insured losses. Drawing on cases such as the 2025 Los Angeles wildfires and Typhoon Hinnamnor, it argues that conventional insurance market solutions are becoming less sufficient as climate risk intensifies, owing to market failure manifested in shrinking insurance supply, growing loss correlations, constraints on insurability, reduced reliability of historical loss data, and the tension between solvency regulation and adequate risk coverage. The paper identifies both demand-side and supply-side drivers of the protection gap and proposes three broad response strategies: strengthening underwriting capacity through data-driven risk assessment, artificial intelligence, climate modeling, and ecosystem building among insurers, reinsurers, brokers, model vendors, and research institutions; expanding non-insurance risk transfer through capital-market instruments, pooled risk-sharing arrangements, and public–private partnerships; and reducing losses themselves through stronger risk awareness, incentives for preventive action, and effective risk communication and evacuation systems. It concludes that because climate risk is cumulative and long term, insurers must establish medium- to long-term climate risk management frameworks, enhance climate-data capabilities and inter-institutional data collaboration, and develop preventive, pre-loss services, thereby enabling insurance to move beyond post-disaster compensation toward strengthening social resilience and sustainability under climate change, with implications for life, health, and long-term care insurance as well.

This paper analyzes the rise of Banking-as-a-Service (BaaS) as a key enabler of embedded, everyday finance in the context of mobile-first distribution and open-API regulation (e.g., PSD2 and the Consumer Data Right). BaaS is defined as the provision of bank-licensed infrastructure and functionalities via APIs, allowing third parties to assemble and deliver financial services; it is differentiated from open banking (primarily data sharing) and neobanks (bank-led digital transformation). The paper outlines the industry and regulatory drivers of BaaS adoption, emphasizing banks’ strategic use of partnerships to extend service reach under constraints such as compliance, privacy, and legacy systems. It characterizes prevailing implementation models (B2B and B2B2C) and the required technical stack, including API gateways, developer portals, and sandboxes, and highlights the need for API-centric core banking architectures and strengthened identity, access control, and data-security capabilities. Comparative case evidence from major international providers illustrates the expansion of BaaS use cases across payments, card issuance, credit assessment, lending, and digital identity. The paper further discusses the Korean market, where service diversification remains constrained but is expanding through bank-led open-API platforms and embedded partnerships with large consumer platforms. Finally, it proposes a forward-looking trajectory toward hyper-personalized embedded finance supported by digital twins and IoT-enabled “zero interfaces,” with implications for proactive financial services and risk management.

○ Since the onset of the pandemic, the accelerated digital transformation and expansion of LLM-based cyberattacks have significantly increased cyber risk exposure across all industries. ○ Recent large-scale personal data breaches demonstrate that security failures at companies with de facto 'societal infrastructure status', such as those in the information and communications technology (ICT) sector and digital platforms, spread across all industries, finance and society. This represents a new form of 'systemic cyber risk'. ○ Big Tech and large digital platforms possess systemically important technologies; their information security failures could thus cause structural shocks similar to systemic risks in financial systems. ○ While the systemic nature of cyber risks has expanded global demand for cyber insurance, the insurance industry's underwriting capacity is being hampered by increasing cumulative risks and state-sponsored attacks. ○ In Korea, while fines for personal data breaches are substantial, liability compensation amounts are very low, limiting corporate civil liability risks. This creates a structural problem where sufficient incentives for cyber insurance adoption are not formed. ○ To address systemic cyber risks, companies must strengthen enterprise-wide risk management systems, insurers must secure security and underwriting expertise, and the government must establish policy foundations such as disclosure requirements, punitive damages, and public-private cooperative insurance programs. ○ Financial authorities must introduce ‘cyber risk stress tests’ based on extreme cyber incident scenarios. This is necessary not only to assess the systemic vulnerabilities of financial institutions but also to quantitatively manage the impact of cyber incidents at big tech and platform companies on financial stability.

Concerns are mounting regarding systemic cyber risks that are expanding beyond the confines of individual companies to encompass entire systems. This phenomenon is precipitated by the rapid advancements in digital transformation and the widespread dissemination of artificial intelligence. This study employed a SEIR model to analyze the chain reaction of cyberattacks and empirically identify the impact of resulting unreported incurred but not reported (IBNR) losses on insurers' solvency. An analysis of 232 ransomware incidents in the financial and IT sectors revealed that higher corporate resilience curbs total losses, while shorter incident reporting delays enable more stable loss management. It is noteworthy that insurers experienced a precipitous decline in their solvency ratio when resilience was low and dependence on the financial sector was high. This impact was more pronounced among small and medium-sized insurers. This study posits that prevailing insurance regulatory frameworks, such as Solvency II, are inadequate in adequately reflecting the distinctive characteristics of cyber risk. To address this, the proposal entails the incorporation of policyholders' security investment levels and recovery capabilities into the underwriting process. The report further proposes the establishment of standards within the prevailing solvency regulation system that reflect the unique characteristics of cyber risk. The implementation of these standards would facilitate the management of reporting delay effects and fat-tail risks. The report further proposes the implementation of a mutual reinsurance system, with the objective of safeguarding small insurers undertaking cyber risk underwriting activities.

This paper examines the escalating threat of cyber risk amid accelerating digital transformation and the widespread adoption of AI, framing it as a prominent form of emerging risk. It outlines the unique features of cyber risk—rapidly evolving threats, high systemic interconnectivity, and challenges in risk quantification—and reviews global market trends, noting the maturity of U.S. and European cyber insurance markets compared to Korea’s underdeveloped landscape. The study proposes a comprehensive framework for Korea, including standardized definitions and classification systems, integrated internal–external cyber risk databases, data-driven assessment models incorporating firmographics and quantitative security metrics, and the development of specialized cyber risk assessment institutions. It emphasizes the necessity of public–private collaboration, regulatory support, and ecosystem building to enhance resilience, improve underwriting accuracy, and foster sustainable growth in the domestic cyber insurance market.

This article analyzes global cyber risk trends and the evolving cyber insurance market, emphasizing the rapid increase in cyber risk exposure since the COVID-19 pandemic, driven by digital transformation and AI adoption. It details the structural characteristics of cyber risks—high interconnectivity, accumulation potential, and difficulty in quantification—and compares the responses of mature markets like the U.S. and EU with Korea’s relatively nascent market. The article recommends establishing standardized risk definitions and classification systems, expanding and integrating internal/external cyber incident databases, developing advanced risk assessment and pricing models, and fostering specialized assessment institutions. It concludes that coordinated public–private partnerships, regulatory frameworks, and improved market infrastructure are essential to enhance resilience, enable effective underwriting, and stimulate the sustainable growth of Korea’s cyber insurance market.