Being an AI-First Insurer in 2028
What will a day of work of an insurer look like in 2028?
That’s not easy to say…
But what is sure is that in 2028, AIs (Artificial Intelligences) will help, will assist — and literally augment — insurers.
And we are not talking about incremental AIs, added on top of existing legacy processes. We are talking about mission-critical AIs, operating at the core of insurers.
AI will be a game-changer. In the future, outstanding insurers will be tech companies before anything else. This vision is exactly what drives the way we build our product and conduct our roadmap at Zelros.
So — what will look like an insurer in 2028? Here are 3 incarnations of the insurance disruption.
Machine Learning scores will replace business rules
Current insurers almost only manage their processes with basic if / then hand-crafted business rules, built from decades of operating experiences:
- if this keyword appears in an incoming email, then dispatch it to this processing team
- if a client has more than X claims in a year, then forward the case to this dedicated person
- if the client is in this age range, and has a credit score in that range, then make a quote with this price
In 2028, almost all decisions will be taken based on machine learning scores, instead of simple, explicit rules. These predictive scores are able to capture much more subtle phenomenons, and are proved to be more accurate than human rules.
Use of machine learning at scale will be possible thanks to availability of rich unaggregated data, and new technologies like AutoML and interpretable AI.
Client centric AI-ready organisations will replace silos
Incumbents carriers are organised in a very sequential way, reflecting the insurance life cycle: product engineering, pricing, marketing, distribution, sales, underwriting, claim processing, fraud detection, churn tracking, customer assistance, …
Each of these domains work in silos, without having in the radar what is at the intersection of all of them: the client.
Insurers of tomorrow will radically change this way of working. In particular, their IT infrastructure will be organised around client-centric, AI-ready data backbones.
For example, when talking to a customer on the phone, a claim manager will not only be able to have contextualised access to algorithmic data related to its perimeter (e.g.: complexity score of the claim), but also from other domains (e.g.: marketing scores like churn or cross-sell).
Insurer-client touch points are rare. AI-first insurers will make each of them an opportunity, by allowing their employees to go beyond their domain expertise tasks, and meet — and even exceed — customer expectations.
Conversations with AIs will replace clicks
Employees of traditional insurers use a lot of software with inconvenient web interfaces, outdated intranet and complex search engines.
For them, interacting with decision making data to serve their clients is a real pain — especially since insurance products are becoming more and more complex. They struggle and lose a lot of time.
The coming years will see the end of the search era, and the advent of assistance. Employees will be augmented with virtual intelligent assistants, that will recommend them the next best actions, and make them more effective.
These AI assistants will be mainly conversational, making natural language the default interface with the insurer core system.
In short, insurance employees will transition from current
“open intranet -> log in -> query -> filter -> read the result -> hopefully find an answer”
to a more efficient and natural way of taking decisions:
“Hey, which plan should I propose to this customer?”
“Hey, what is the status of M. X reimbursement?”
“Hey, what is the minimum refund amount of this plan?”
Towards end-to-end AI-operated insurers
This will be the path towards end-to-end AI-operated insurers, delivering new and better client services — and we at Zelros are on a mission to be at the center of this revolution.