Insurers still miss customer expectations despite years of modernization

The expensive mistake in insurance tech is assuming that another model will fix a broken journey. Carriers have spent years upgrading pricing, underwriting, and digital distribution, but policyholders still run into stale claim updates, slow service responses, and confusing communications because the systems behind those decisions are not working as one.

Modernization did not equal orchestration

Earnix’s core point is uncomfortable precisely because it is so practical: the industry’s problem is not a lack of technology spend. It is the gap between internal decision systems and the customer-facing experience those systems are supposed to improve. If data, pricing logic, underwriting rules, and digital channels are not orchestrated together, the insurer can become more automated without becoming more responsive.

That distinction matters in both personal lines and commercial lines. Customers now expect faster quotes, clearer offers, and more transparent outcomes across the lifecycle, not just a slicker website or a better sales flow. When carriers treat customer experience as an add-on instead of an operating design problem, the policyholder feels every handoff, every delay, and every disconnected update.

The numbers show broad adoption, but thin delivery

Earnix’s 2026 Insurance Trends Report, based on a survey of 400 global insurance executives, shows just how widespread AI adoption has become inside insurance organizations. According to the report, 81% of executives say AI is embedded across most or some workflows, and 80% are experimenting with or planning to adopt generative AI within the next two years.

But the same research shows how uneven that adoption still is in the places customers notice most. Fewer than a quarter of insurers currently use AI for claims processing, at 23%, policy issuance at 18%, or churn prediction at 15%. Even more telling, 56% of insurers prefer a gradual GenAI approach that keeps human intervention in place for at least the next three years.

That is not a story about a lack of interest. It is a story about caution, fragmentation, and an industry still deciding where AI belongs. The strongest evidence of progress is that the tools are inside the workflow. The strongest evidence of weakness is that the tools are still not shaping enough of the customer journey to feel visible.

The bottlenecks are still in the plumbing

WTW’s 2024 P&C Insurance Advanced Analytics Survey puts the execution problem into sharper focus. It found that 49% of insurers were incorporating AI into their analytics processes, but IT bottlenecks were the most frequent barrier to broader use. Progress had slowed even after significant investment, which is usually what happens when the software stack is the real constraint instead of the algorithm.

That is the part many carriers still underestimate. Legacy core systems can make a promising model look good in a pilot and weak in production. Disconnected CX workflows can leave claims, policy service, and underwriting operating like separate businesses. Weak orchestration between AI models and frontline servicing means the customer gets automation in one place and a manual workaround in another.

This is why adding another front end often fails to move the needle. If a carrier modernizes the pricing engine but leaves claims updates buried in a different workflow, the customer sees inconsistency, not innovation. If underwriting rules improve but the service desk does not have a clean handoff, the policyholder still has to chase answers.

Front-office AI is rising, but trust is fragile

EY’s GenAI survey shows that insurers are trying to push AI closer to the customer. It found that 56% of insurers are prioritizing front-office GenAI use cases, and 63% of P&C carriers, life and annuity carriers, group benefits providers, and brokers are making similar investments. That is a clear signal that the industry knows where the visible value should be.

The problem is that consumers are not automatically buying the pitch. Insurity’s 2025 survey found that only 20% of Americans said it was a good idea for P&C insurers to leverage AI, down from 29% in 2024. The same study found that 44% of consumers were less likely to buy a policy from an insurer that publicly uses AI, and positive consumer experiences with AI fell from 63% to 47% year over year.

That gap matters because public trust is now part of the product. Customers may never see the model behind an underwriting decision, but they will absolutely notice whether the claim update arrives on time, whether the service response makes sense, and whether the company can explain what happened without making them repeat the same story three times. When AI is visible but not helpful, it stops looking like progress.

What carriers need to optimize next

The practical lesson for carriers is straightforward: AI spend only becomes customer value when the operating model changes with it. Better segmentation, more relevant offers, faster decisions, and fewer handoffs are the real goals. Anything less leaves the insurer with more automation and the same old friction.

The software buyers getting this right are looking for platforms that connect pricing, underwriting, claims, and customer communication instead of optimizing each layer in isolation. The test is not whether a model scores well in a lab. The test is whether a policyholder gets a clearer answer, a faster update, or a cleaner next step.

That is why this story is not about AI hype. It is about execution. Insurers already know how to buy technology; the harder job is making sure the technology changes what the customer feels. Until pricing, service, and communication move together, modernization will keep producing impressive internal dashboards and underwhelming policyholder experiences.