.png "Claims Management in Augmented Underwriting: Adapting to New Underwriting Models")
.png "Claims Management in Augmented Underwriting: Adapting to New Underwriting Models")
Introduction
The insurance industry is continually evolving, driven by advances in technology and data which is enabling shifts in underwriting models.
In the Q4 2024 Oxbow report commissioned by the LMA, several new methods of underwriting were described:
In light of this research, it is equally important to consider how the claims function will similarly evolve and how it will potentially reshape how claims are assessed, managed, and resolved, with implications for efficiency, accuracy, and the role of the human-in-the-loop.
The focus of the report on enhanced underwriting models which are described as “propositions that use data and digital technology to enhance underwriting decisions, and propositions that have taken a new strategic approach to follow business”. These models leverage algorithms to enhance human decision-making, codifying and quantifying risks.
Claims management in the age of augmented underwriting
Augmented underwriting which applies to both lead and follow underwriting, introduces varying levels of system rather than human judgement i.e. from simple rules-based triage all the way to sophisticated risk scoring and insights using data and/ or technology. Some further still which assess propensity to go to litigation for example.
Aside from underwriting, claims teams are similarly using data and technology for automation via Agent AI, Argentic AI as well as using AI for tasks that a humans simply could not do such as data insights across large volumes of documents.
For claims teams therefore, the addition of augmented underwriting alongside claims’ own digitisation journey should mean greater data collaboration between underwriting and claims functions. The now easier availability of data that supports enhanced underwriting will allow Claims Handlers/ Adjusters to access more granular risk insights than today, ideally leading to faster and more accurate claims decisions. This, coupled with Claims’ own use of data and AI, will mean claims teams can focus on high-complexity claims where human judgement is critical and where routine, low-value claims are processed more automatically.
But… there is the inevitable challenge of adoption and implementation of these data-driven insights into claims processes. Claims teams will need to develop expertise in data and interpreting algorithm-generated data, and claims teams will need to increasingly rely on structured data and real-time insights generated at the underwriting stage.
With algorithmic underwriting, where there is 100% automated decision-making based on predefined rules and risk codification, ideally allowing for straight-through processing (STP), similarly, the claims function needs to consider what this means for a lead or follow position.
1. 100% Line Model
Where risks are underwritten 100% by an algorithm, claims processes need to similarly become more standardised and aligned to the rules created by the algorithm, and to minimise differences in interpretation and policy understanding. Claims teams will need to understand this automation, their own rules and standardisation and where human oversight is needed for exceptions and disputes.
2. Syndicated Follow Model
The Oxbow report calls out two key implications for carriers of these new models:
• Squeeze on traditional open market follow: As leaders take a larger share, follow capacity may shrink.
For claims, this will also consequently reduce the number of claims agreement parties potentially leading to faster decisions as fewer carriers/ TPAs are involved. The role of the second lead/ first follower may also evolve in their role as a second pair of eyes on the lead’s decisions.
• Increased reliance on lead carriers for risk assessment: Leader fees could be introduced to reflect the added workload.
For claims, there will be a higher demand for expertise for the assessment of higher complexity claims (as the routine is automated). In a market where such talent is reducing, could this then offset the potential for faster claims settlements as more claims are triaged to fewer available experts? And how do we develop this specialist expertise with this increased automation?
Implications of digital and algorithmic broker facilities on Claims
Another model that has been around for some years, albeit now becoming more data-led and automated with the use of algorithms and AI, is broker facilities. This model is now underpinned by a flow of data via APIs to carriers and offers real-time data analysis and when algorithms are used, carriers can adjust their risk appetite much more quickly. Given the nature of some of these facilities however, across the ‘good and the bad’ of a book of business, overall, there might reasonably be a greater number of claims to handle. Where a portfolio contains every/ many lines of business, if some of these unfamiliar lines of business, similarly, the claims expertise may not be strong.
Active portfolio trackers
Active portfolio trackers are mostly focused on a more effective way to deploy capacity and manage risk with a focus on assessing, and processing risks efficiently. Whilst this delegation significantly reduces admin costs, the claims function still needs to be considered. Similar and different to facilities, admin efficiency gains should be felt by the claims teams.
For claims teams, the benefits of these facilities and tracker models include:
- Improved data accessibility: Real-time data and via API means claims handlers can access and analyse data much more quickly leading to faster response times. The fact that data is structured also enables wider analysis of trends and fraud detection.
- Operational efficiency: Digital and structured data reduces manual input and/ or rekeying.
However, a word of caution, as these broker-led facilities are digital in nature with structured data, claims systems and processes must also be able to work in the same way and be set up to create the analytics to get the most value and efficiency from this model.
It’s fair to say that the future of claims is of course interlinked with the evolution of underwriting models. For claims teams, these models, as with any change, offers both challenges and opportunities. The key is to become more digital and data-savvy in order to access the consequent efficiency gains, and data-driven insights whether these models increase or decrease the volume of claims to handle. It is clear however that the routine and simple claims will be more automated with claims talent needing to focus on the more complex and themselves be data-savvy. Again, the question of talent and talent development must be addressed sooner rather than later.
Irrespective of the model(s), the claims function is still ‘what we sell’ and should hold that importance and focus as underwriting evolves.