AI pushes insurers toward proactive, personalized customer service

The new AI pitch in insurance is not just automation

At Insurance Innovators USA in Nashville, the conversation was not about shaving a few minutes off a call. It was about making service feel like a concierge experience, where an insurer anticipates what a policyholder needs before the customer has to repeat the story. That shift matters because the event, held May 11-12 at Music City Center, drew more than 1,500 senior insurance professionals from carrier and software ranks, and the tone was unmistakable: AI is moving from back-office efficiency toward front-line customer design.

Russell Page, chief information officer at Hagerty, captured the mood with a simple idea: personalization should feel like a concierge service. That sounds polished on stage, but in practice it sets a much higher bar than a chat widget or a few automated email triggers. It means the carrier has to recognize context, route the right task, and make the interaction feel intentional across billing, first notice of loss, renewals, and everyday service requests.

Why the service problem is now a retention problem

The push for hyper-personalized service is happening because insurance buyers have become less forgiving. J.D. Power’s 2025 U.S. Insurance Digital Experience Study found that 57% of auto insurance customers actively shopped for a new policy in the past year, the highest shopping rate it had recorded. That is a sharp warning sign for carriers that still treat service as a cost center instead of a retention engine.

The same pressure shows up in claims. J.D. Power said 52% of auto and homeowners customers who rated their digital claims experience as poor or just OK were likely to leave or not renew. That is the operational tradeoff behind all the AI talk: if the experience is clumsy, digital does not just fail to delight, it actively increases churn risk. In that environment, better routing and faster response are not nice-to-haves. They are part of keeping the policy in force.

What AI can actually improve across the policy lifecycle

The most useful way to think about AI in P&C is by workflow, not by feature. Billing is an obvious starting point because payment questions, installment changes, and confusion over notices generate a steady stream of service contacts. If AI can identify the reason for the contact early, pull the right account details, and hand off a ready-to-act summary to a human or self-service flow, the customer gets a cleaner answer and the contact center avoids a dead-end call.

FNOL is even more sensitive. The first notice of loss is where confidence can disappear fast, and it is where generic automation usually falls apart. A good AI layer should help a carrier understand the claim context, surface the next best step, and reduce the number of times a customer has to restate the same facts. Renewals benefit in a different way: AI can flag policyholders who may need outreach before the renewal date, especially if recent service history, billing friction, or claims activity suggests a higher leave risk.

That is the real appeal of proactive service. It is not just about answering faster. It is about spotting where a policyholder is likely to stumble and getting ahead of the stumble with the right message, the right channel, and the right person.

The software stack behind the promise is the hard part

This is where the front-end story often outruns the back end. A slick AI assistant is easy to demo, but hyper-personalized service depends on whether the carrier can connect CRM, policy administration, claims, billing, and customer data in a way that actually works in production. If the contact center sees one version of the customer, the policy system sees another, and the renewal engine is working off stale data, the personalization layer becomes another disconnected screen.

Page’s role at Hagerty shows why this is as much an infrastructure problem as a customer-experience one. Hagerty says he leads IT strategy covering analytics, data science, cyber and information security, corporate systems, and network and infrastructure management, while also safeguarding customer profiles. That is the kind of remit you give someone when the job is not merely to deploy tools, but to make sure the tools can trust the data underneath them.

A customer data platform can help unify signals, but it does not replace the policy system. CRM can organize interactions, but it does not by itself know whether a billing issue, a claim, or a renewal is the real trigger for outreach. The carriers that get this right will build a service layer that can read across the lifecycle, not just respond to the latest inbound message.

Hagerty is a good case study because context matters

Hagerty is not trying to serve the average mass-market customer. The company describes itself as a specialty insurer focused on the global automotive enthusiast market, including classic cars and other enthusiast vehicles. That kind of business depends on context-rich relationships, where the insurer often knows the customer’s vehicle, usage pattern, and expectations better than a commodity carrier would.

That makes Page’s comment about concierge-style personalization especially telling. In a niche like Hagerty’s, a generic response is not just inefficient, it can feel tone-deaf. The service model has to recognize that the value of the policy is tied to more than price. It is tied to expertise, memory, and the sense that the carrier understands what the customer cares about.

Trust still decides whether AI helps or hurts

There is a reason insurers cannot just slap AI on top of service and call it progress. Insurity said that a January 2025 survey of more than 1,000 U.S. adults found declining consumer confidence in AI for P&C insurance. That is a real warning for any carrier assuming customers will automatically welcome more machine-driven interaction. If the logic is hidden or the outcome feels opaque, the personalization layer can quickly become a trust problem.

IDC’s view adds another piece to the puzzle. The firm said AI adoption in insurance is shifting from productivity-focused use cases to strategic growth drivers, with P&C insurers using AI to improve customer experience and operational efficiency while still facing data-integration and trust challenges. That is the right framing. The goal is not to automate everything for its own sake. The goal is to make service more responsive without stripping away the judgment and accountability that insurance still requires.

The carriers that will win this phase are the ones that treat AI as service orchestration, not just a chatbot project. If the technology can pull billing, FNOL, renewal, and service data into one coherent operating model, it can cut friction, lower call-center volume, and reduce retention risk. If it cannot, it will just add another layer between the customer and the answer.