Generative AI and the Insurance Customer Experience

Generative artificial intelligence (AI) has been a game-changer for many industries, and insurance is no exception. From policy selection to claims settlement, generative AI is rewriting the rules of how customers interact with insurers throughout the insurance customer journey. On one hand, insurers are deploying generative AI tools to enhance customer interactions; on the other, customers are independently using general-purpose, off-the-shelf generative AI apps to inform their own insurance decisions. This two-way application of generative AI is transforming experiences at each stage of the journey, raising both new opportunities and new questions about trust, transparency and value.¹

The insurance customer journey typically consists of five stages: 1) gathering information, 2) seeking advice, 3) purchasing a policy, 4) receiving after-sales service and 5) filing claims. We believe generative AI has the potential to elevate each of these stages by providing faster, more personalized and more efficient interactions than ever before.

At the same time, insurers and actuaries are no doubt busy navigating the challenges associated with these enhancements, ensuring that the human element, data privacy and fairness remain front and center.

Here, we explore generative AI’s dual role in insurance, how it impacts each step of the customer journey, what customers are saying about these changes and how industry professionals can respond strategically to harness generative AI’s benefits while managing its risks. Much of the information summarized in this article is drawn from a recently published Geneva Association report, “Generative AI in the Insurance Customer Journey,” as well as the accompanying global survey of insurance customers across the six largest insurance markets.

Generative AI’s dual insurance role

Generative AI can be thought of as playing two distinct roles in the insurance ecosystem.

First, there are insurer-provided generative AI tools, which include chatbots, virtual assistants, AI-driven underwriting and claims algorithms, and other systems that insurance companies use to serve and communicate with customers. These tools can handle routine inquiries, generate quotes or policy documents, personalize recommendations and even help detect fraud. They act as an extension of the insurer’s workforce—a “digital concierge” available 24/7 to provide information or assistance. For instance, Ping An Insurance in China uses a generative AI chatbot to handle millions of customer queries and assist with drafting policy and claim documents, even providing financial advice in natural language. By deploying such tools, insurers can potentially improve efficiency, service consistency and, importantly, the customer experience.

Second, there are off-the-shelf generative AI platforms used by customers on their own initiative. These include general-purpose AI chatbots and apps or services that consumers increasingly turn to for independent research and advice on insurance matters. Customers are using these tools to analyze insurance products and clarify coverage questions on their own, so they often arrive better informed—and with higher expectations—when they engage with an insurer. In effect, generative AI is empowering customers to take more control of their insurance decisions, making them more proactive participants in the process. Generative AI is empowering customers to take more control of their insurance decisions, making them more proactive participants in the process.

This dual role of generative AI is reshaping the insurer-customer relationship. We believe insurers that enhance their own AI capabilities while embracing customers’ independent AI use are likely to deliver better customer experiences.

Impact across the customer journey

Every step in the insurance customer journey can be affected by generative AI. This is how we see generative AI impacting five key stages:

1. Information and research: In the initial stage of customers gathering information about risks and insurance options, generative AI expands access to knowledge. A prospective buyer can ask an AI assistant to explain the nuances between two insurance policies and get an instant, tailored answer—potentially creating a more informed customer from the outset. Generative AI tools also simplify complex insurance terminology into plain language, making it easier for consumers to understand coverage details.
2. Advice and recommendations: When seeking advice on what insurance to buy, generative AI provides a new level of personalized guidance. Generative AI can analyze an individual’s profile and needs against thousands of products in real time to suggest suitable options – a breadth of analysis beyond what any single human broker could manage. Unlike earlier robo-advisors that offered one-size-fits-all suggestions, generative AI can have a nuanced dialogue about trade-offs, explaining why one policy might be better suited than another for the customer’s particular situation.
3. Purchasing and underwriting: We’ve observed that generative AI is beginning to streamline the process of purchasing a policy and underwriting risk. Generative AI systems can dynamically assemble personalized coverage packages and instantly evaluate data (from apps, medical records, Internet of Things devices, etc.) to price policies, flagging only unusual cases for human review. Some insurers are even testing conversational interfaces where a customer can “talk to” a generative AI chatbot to complete a purchase end-to-end. While complex or high-value policies still require human oversight, generative AI is steadily taking on more of the workload in this stage.
4. After-sales service: Once a policy is in force, generative AI-powered virtual assistants are elevating customer service. Unlike basic chatbots of the past, modern AI agents can handle a wide range of service requests and questions around the clock, often without human intervention. For example, an AI service agent might guide a customer through updating their address or adding a beneficiary via a simple conversation. And because today’s generative AI can process text, voice and even images in one unified system, customers usually get seamless, consistent answers on their preferred channel. However, for complex or emotionally sensitive issues, most customers still want the option to talk to a human, so generative AI should assist but not completely replace human service representatives.
5. Claims handling: The claims process—or, as we call it, the “moment of truth” in insurance—is also accelerating with generative AI. AI models can analyze photos of damage or interpret medical reports to help adjusters make quicker, more informed decisions. In straightforward cases, it’s conceivable that a customer could report a loss to an AI chatbot, upload evidence and receive a near-instant payout offer if everything checks out. But if generative AI makes a mistake or denies a valid claim, it can undermine trust. For this reason, many insurers use generative AI to support human adjusters rather than replace them; routine claims might be settled with minimal human input, but complex claims still get a human review. This hybrid approach provides speed where possible while ensuring empathy and oversight remain in the loop.

Overall, the infusion of generative AI at each stage of the journey augments the capabilities of both customers and insurers. Customers get more information and faster service; insurers gain efficiency and can personalize offerings at scale. These benefits, however, depend on carefully managing the challenges introduced by generative AI – especially around trust and accuracy – as the next section discusses.

Customer attitudes and concerns

How do customers feel about this AI-driven transformation of their insurance experience? A 2025 Geneva Association survey of insurance customers found attitudes that are cautiously positive overall. The survey was conducted in February 2025 across the six largest insurance markets worldwide: the U.S., China, Japan, the U.K., France and Germany. In each market, 1,000 individual insurance customers were surveyed; samples were designed to be representative of the insurance customer profiles in their respective markets.

People—especially in tech-forward markets like China and the United States—see clear benefits to AI-driven insurance services. They appreciate the prospect of faster service, 24/7 availability and more tailored products. However, they also have clear conditions for embracing generative AI in insurance.²

Figure 1: Customer priorities around generative AI and insurance

Source: Insurance customer survey, the Geneva Association (November 2025)

Four key expectations emerged from the survey:

1. Data security: Security is paramount. Customers need assurance that their personal data is safe and won’t be misused by AI systems.
2. Accuracy: Customers expect AI-driven advice and decisions to be correct. Knowing that AI can sometimes sound confident but be wrong, they want insurers to double-check AI outputs and ensure the accuracy of what is being presented to them.
3. Human backup: Customers insist on being able to reach a human when needed. They’re fine with generative AI handling routine tasks, but for complex or sensitive issues, they want a real person available; without that option, many could lose trust.
4. Transparency: Customers want transparency about AI’s use. This means knowing when they’re interacting with AI versus a human, and getting clear explanations for any AI-made decisions (for example, understanding why AI recommended a certain product or why generative AI declined a claim).

The survey also shows that acceptance of fully AI-automated services drops off for more complex tasks. Many people are uneasy with the idea of an entirely AI-driven process for major decisions – for instance, buying a complex life policy or handling a large claim – without any human oversight. In practice, customers might trust generative AI’s initial answer, but they still want a human to confirm or be available for critical outcomes.

For insurers and actuaries, the message is clear to us: Any generative AI initiative should be designed with customer preferences in mind. Innovations that fail to offer human support, or that compromise privacy, accuracy or transparency, will likely face customer pushback regardless of their technical prowess.

Key takeaways for insurers and actuaries

To harness generative AI’s potential while maintaining customer trust, insurers (and the actuaries guiding them) would ideally welcome four strategic steps, in our opinion:

1. welcoming the knowledgeable customer empowered by generative AI
2. maintaining human oversight
3. putting solid guardrails around generative AI
4. safeguarding data.

The common thread is aligning technological capabilities with the core values of insurance (trust, fairness, customer-centricity) so that innovation comes with reassurance.

One of the most insurance-specific and forward-looking insights from the consumer survey lies in the emerging dual pathway through which customers interact with generative AI: via insurer-provided tools and off-the-shelf apps and platforms. For insurers, the growing use of third-party generative AI by consumers introduces a parallel advisory ecosystem beyond the company’s control—one where customers seek policy advice, compare products or interpret terms using general-purpose generative AI not trained on insurer-specific data.

This raises challenges not just in terms of misinformation and liability, but also in managing expectations shaped outside the insurer’s domain. In contrast, insurer-provided tools come with greater control and compliance responsibility, but also the opportunity to shape trusted, consistent customer experiences. We believe insurers would do well to develop a dual strategy in turn—one that governs and optimizes internal generative AI use, and another that acknowledges and supports the generative AI-enabled consumer—helping ensure alignment, transparency and resilience across both interaction modes.

Conclusion

Generative AI is transforming the insurance customer journey, bringing possibilities alongside new challenges. For actuaries and insurance leaders, the goal is to leverage this technology to enhance customer value and efficiency without undermining trust. Companies that proactively embrace generative AI, underpinned by human oversight and ethical principles, could be well-positioned to meet the rising expectations of tomorrow’s insurance consumers.
The promise of generative AI in insurance is substantial: more personalized coverage, faster service, and broader access for consumers. But fulfilling that promise requires navigating the issues of privacy, accuracy and fairness. Actuaries, with their expertise in risk and data, are well-positioned to ensure that AI models are reliable and unbiased, and are there to help interpret and communicate AI-driven results. By updating some traditional practices with new AI-driven approaches—while holding tight to fundamental principles of fairness and transparency—insurers can turn generative AI into a true tool for innovation and customer trust.
In the end, insurance’s mission remains the same: helping people manage uncertainty. Generative AI, used wisely, can become a powerful new means to that end.

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries or the respective authors’ employers.

Copyright © 2025 by the Society of Actuaries, Chicago, Illinois.