The application of artificial intelligence in the insurance industry is evolving from single-point efficiency improvements towards comprehensive business model restructuring.
Recently, Yu Bin, Vice President of China Pacific Insurance (Group) Co., Ltd., stated that the most critical strategic decision for insurance companies is no longer whether to adopt AI, but rather whether to remain in the optimization phase of "AI integration" or advance into the transformation zone of "AI-driven" operations.
According to Yu Bin, the shift from "AI integration" to "AI-driven" represents more than just deploying additional models or implementing more use cases; it signifies a fundamental change in mindset and business logic. "AI integration" resembles equipping existing processes with tools to address localized pain points, whereas "AI-driven" approaches treat AI as a core underlying capability to reinvent products, services, processes, and even business models. Consequently, the future competitive focus for insurers may not solely revolve around point-based efficiency gains, but rather around which companies can sooner complete the leap from localized optimization to systemic transformation.
Regarding core insurance operations, Yu Bin highlighted that AI is most likely to drive initial changes across three dimensions: sales, operations, and organizational structure. On the sales front, as customers increasingly utilize tools like Doubao and DeepSeek for insurance product inquiries and comparisons, new traffic channels are altering customer decision-making processes. In the future, insurers may not only compete for customers through traditional channels but also engage in data and product interactions with dominant AI platforms. Those who adapt to this change earlier are more likely to gain a competitive advantage.
In operations, collaboration among AI agents is expected to reshape processes such as policy application, underwriting, claims settlement, and customer service. Yu Bin illustrated this with an example in auto insurance claims: in the future, AI could autonomously coordinate multiple agents to complete tasks, including guiding customers to document accident scenes, invoking image recognition models for damage assessment, matching repair shops and parts information, and推送 settlement options to clients. This perspective suggests insurance operations may evolve beyond mere "human-machine collaboration" for localized efficiency gains towards new workflows driven by multiple intelligent agents.
Looking further ahead, changes are anticipated in the organizational structure itself. Yu Bin indicated that as "human-in-the-loop" feedback mechanisms become established, models could develop continuous learning capabilities, prompting insurance institutions to explore AI-native organizations. Traditional structures organized by business lines, departments, and roles might be reshaped by algorithm-driven, dynamically iterative collaboration models, thereby amplifying employees' productivity leverage.
When discussing the return on investment for AI in core insurance businesses, Yu Bin emphasized the need for a multidimensional evaluation beyond单一 financial metrics. This includes customer relationship management, industry risk mitigation, workforce capability enhancement, operational efficiency improvements, investment quality enhancement, and compliance risk reduction. The value of AI manifests not only in improved customer experience, increased premium per employee, and optimized expense ratios but also in expanded risk coverage, enhanced investment capabilities, and reduced risk losses.
Current progress shows clear effectiveness in certain areas. For instance, in claims processing, CPIC has increased automation in health insurance through AI, reducing average individual claim settlement time to two minutes. In sales, tools like the AI Gold Coach and Customer Management Assistant are upgrading marketing models, boosting new premium income per agent by 21% in pilot regions. In risk control, AI algorithm-based pilots in health and auto insurance have generated over 100 million yuan in additional loss reduction this year.