The Next Phase of Insurance AX: AI Agents In Customer Service
Hello, this is Treasurer.
Imagine calling an insurer to check the status of a claim.
Traditionally, an agent would verify your information, search through policy and claims systems, and then provide an answer.
That model is changing.
Financial institutions are moving beyond simple call automation toward AI systems that can understand customer requests, connect relevant data, and turn each interaction into usable information.
A recent example is KT’s next-generation AI Contact Center (AICC) for NH NongHyup Bank, a project worth approximately KRW 40 billion.
The number of tasks handled directly by AI increased from 45 to 180, speech recognition accuracy improved from around 90% to 97%, and customer wait and handling times fell by roughly 20%.
From 45 Tasks to 180
Source: 백세시대
AICC, or AI Contact Center, applies technologies such as speech recognition, chatbots, and voice bots to customer service.
Earlier systems were largely limited to scripted answers and simple inquiries. NH NongHyup Bank’s new AICC significantly expands the range of tasks AI can complete without transferring customers to a human agent.
The focus is shifting from simply answering calls to how much of the customer request AI can actually resolve.
Connecting Calls, Chat, and Email
Source: Call Centre Helper
The new system also integrates video calls, chat, and email into a single operating platform.
If a customer starts with chat and later calls, disconnected systems may force them to explain the same issue again.
With connected interaction histories, agents can immediately see previous questions and responses.
This is the idea behind an omnichannel experience: different channels, but one continuous customer journey.
Why Data Connectivity Matters More in Insurance
Source: 토큰 포스트
Insurance inquiries often require more than a standard FAQ.
If a customer asks, “Is this surgery covered under my policy?”, the answer may depend on:
- Policy details
- Riders and coverage terms
- Claims history
- Current claim status
- Previous customer interactions
For AI to support insurance service effectively, it must do more than understand language. It must also find and connect the right information at the right time.
How AI Agents Can Support Advisors
Source: 바이라인네트워크
Financial institutions are now moving from traditional chatbots and voice bots toward AI agents.
KT has said it is applying agent-based technology to financial AICC systems and developing generative AI customer-service support for the insurance sector.
In practice, an AI agent could identify a customer’s request, retrieve relevant policy or claim information, and organize the key details for an advisor.
For example, when a customer asks why a claim payment has been delayed, AI could surface the relevant claim and current processing status before the advisor begins searching across multiple systems.
Every Conversation Becomes Data
Source: Zenloop
Another important shift is that customer conversations themselves become valuable data.
Repeated questions can reveal where customers face friction, which policy terms are difficult to understand, or where the claims process may need clearer guidance.
Instead of remaining as archived call records, these interactions can become Voice of Customer (VOC) data that helps improve products and service processes.
The Next AICC Advantage Is More Than Better Answers
Source: Development Beyond Learning
The early question for AI customer service was simple:
“How accurately can AI understand what customers say?”
Now the questions are broader.
- Can AI access the right customer information?
- Can it use previous interaction history?
- Can it connect records across different channels?
- Can those interactions improve future services?
AICC is therefore evolving from an automated response system into a platform that connects customer interactions with business data.
Data Connectivity Is Also Central to Insurance AX
Customer service and reinsurance settlement are different processes, but both depend on connecting information scattered across multiple systems.
Treasurer’s reinsurance AX solution structures and connects contracts, claims, cover notes, and settlement records to support reconciliation workflows.
In customer service, the key connections may be between questions, policies, claims, and interaction history.
In reinsurance, they may be between contracts, claims, cover notes, and settlement records.
In both cases, the value of AI depends not only on the model itself, but on how well business data is structured and connected around it.
Financial customer service is moving beyond callbots toward AI agents. In insurance, the next step is not simply answering more questions, but turning every customer interaction into connected, actionable data.