The process of buying insurance has always been complex. The very first step—deciding to buy life insurance—is considered ‘emotionally heavy’. Then, the buying journey can be daunting, with long proposal forms and multiple questions about lifestyle, personal, and family health histories. Providing relevant documents for KYC, income proof, etc., along with physical medical tests… the list is unending. Even after your form is submitted, underwriters take up to six weeks to analyse and assess the individual’s risk and decide on the policy’s issuance.
However, as artificial intelligence (AI) gets more deeply integrated into the life insurance value chain, we can say goodbye to the complexities for good! Life insurance is transforming; underwriting, pricing, and the claims process will not be the same again.
Tech trends reshaping life insurance
Three core technology trends are reshaping life insurance underwriting and claim processing.
Avalanche of data from connected devices
The number of connected devices (IoT) and signals collected from them are increasing exponentially. The penetration of devices such as smart homes, smart cars, telematics, smart homes, wearables, etc. is increasing, and new categories of devices and sensors such as eyewear, shoes, medical devices, etc. are also added to the list. By 2025, the number of connected devices will reach an astronomical one trillion mark.
The vast amount of data collected through these connected sensors is enabling insurers to understand customers deeply and assess risk more accurately at the individual level to offer personalised products and pricing.
For example, traditionally life insurance has been priced considering some broad parameters such as age, gender, and smoking habits. These parameters make up a cohort. These cohorts often overlook other granular details that can give us more information about the customer’s health, lifestyle, and different financial needs.
If there is an adverse claim experience in such a broad-level cohort, it impacts the pricing for all individuals in the cohort, no matter how different they are.
Data from wearables can tell us about an individual’s physical activity and health status and can determine the price of the insurance product more effectively.
In the age of IoT, customers can ‘pay as s/he lives.’ Connected devices also allow insurers to engage with the customers continuously and offer new services and products, a shift from a business model in which touchpoints are limited to annual renewals and occasional claims.
Emergence of shared data-ecosystem:
Various public and private entities are coming together to create an ecosystem to share customer data, to enhance customer experience. With customer consent, this alternate data available to insurers can be used to assess underwriting risk in real time to accelerate issuance or validate claim information to reduce the turnaround time for claim approvals.
As customers own the data, if the policy is shifted from one insurer to another, the data and the risk profile of the customer can also be ported. The customer’s information is available through IIB (India Insurance Bureau), AA (Account Aggregator), KYC, etc., which are good examples. Initiatives such as Bima Sugam will further accelerate the progress in the shared data ecosystem.
Advancements in cognitive computing:
As the cost of computing has drastically come down, and more and more data is available, the algorithms are constantly learning and adapting to the world around them. Cognitive computing, or deep-learning models, is an imitation of human neurons, which is based on a human’s ability to learn through decomposition and inferences.
With the increased commercialization of these pre-trained models and plug-and-play platforms the insurers can easily access these models at nominal cost and further augment them with the customer data available with them.
Redefining Underwriting
These technology advancements are transforming the fundamentals of underwriting.
From Risk Mitigator to Risk Preventor: Traditionally, insurers got involved only after a claim was raised. Their main objective was to minimize the incurred losses, playing the role of a ‘risk mitigator’. With the availability of the vast amounts of data available for continuous risk assessment, proactive measurement of risk can be done. Before the damage occurs, it can be prevented in certain situations. For example, data available through wearable and medical devices are enabling insurers to assess health risks continuously and proactively engage customers to prevent damage in case any adverse signal, such as fall risk, cardiac arrest risk, etc., is detected.
‘Traditional Data’ to ‘Alternate Data’ based underwriting: To assess risk linked with a customer to determine issuance eligibility, underwriters require information such as intent to buy, income proofs, and medical records. Traditionally, this information is provided by the customers in the form of documents and medical test reports. In the age of data and AI, the risk assessment can be done using alternate data such as customers’ identity can be directly validated through Government databases such as NSDL, cKYC etc., instead of asking for income documents to determine the eligible cover, financial surrogates such as Loan details available through Bureau, Vehicle details available through Vahan API etc. can be used.
Similarly, other behavioural signals and digital footprints based on device, IP address, location, etc., can be analyzed to filter out fraudsters. Determining health risks using alternate data is challenging, but using deep-learning algorithms, lifestyle diseases such as cardiovascular disease and diabetes can be predicted with high accuracy based on information available on lifestyle; these health predictions combined with vital signs extracted using facial analytics can be used as health surrogates instead of physical medical test. Health data available through ABHA (Ayushman Bharat) eco-system will further boost alternate data-based risk profiling and algorithmic decisions.
‘Cohort level’ to ‘Individual Level’ Underwriting and Pricing: Traditional actuarial pricing models are at customer cohort levels such as age, gender, and smoking because of the unavailability of more granular mortality tables. Digital technologies, Data and AI are enabling insurers to assess risk at individual level, and price the product more accurately as per the individual’s risk profile, behaviour and lifestyle.
For example, health history combined with physical activities data available through wearable can be a good measurement of health risk at an individual level, and can be used to appropriately price the premium, to ‘pay as they live’.
‘Manual’ to ‘Algorithmic automated real time’ underwriting: Traditionally, underwriters, after collecting all the information from customers, use an ‘underwriting manual’ available in the form of a software bench to assess an application to make subjective issuance decisions. The process is manual and involves subjectivity. The decision can vary from one underwriter to another based on their individual biases and subjectivity. This manual decisioning is getting transformed using AI. Cognitive algorithms can identify the hidden patterns in risk indicators available through alternate data-ecosystem, and risk scores a customer in real-time. As algorithms can continuously learn themselves, over time, the accuracy of issuance decisions taken by algorithms will supersede the manual underwriters’ decisions, reducing false acceptances and rejections. As the accuracy of underwriting decisions taken by algorithms will improve, the claims events will reduce, further reducing the premium cost, making insurance faster and more affordable for the ‘good profiles’.
‘Annual’ to ‘Continuous’ risk assessment: Data and AI are driving the transition from ‘purchase & renewal’ to a continuous cycle, a product that is continuously adapted to an individual’s behavioral pattern. Life Insurance, in which touchpoints have been limited to annual renewal and occasional claims events, data available through alternate data-ecosystem and connected IoT devices enables insurers to continuously engage with customer for continuous risk assessment to prevent damage, and service better.
Redefining Insurance Claims Process
The evolution of Data and AI technologies, analytics capabilities and customer preferences are also transforming the way insurers manage and process claims.
i. Human and AI team up to detect fraud
a. Events of claims are the fundamental reasons why insurers exist. Intimation of a claim is not a problem, but fraudulent claims are. According to industry estimates, insurers lose ~10% of the premium collected to fraud. Insurers will continue to combine and harness the best features of AI and human intelligence to detect fraudulent claims and expedite the processing of genuine claims.
b. Digitally enabled claims handlers and algorithms will continue working and complementing each other to expedite claims processing. While algorithm decisioning and data science will augment the claims handler’s decision to accelerate the processing, the claims handler’s robust knowledge of claims patterns, fraudsters’ behaviors, and causes of losses will strengthen algorithms.
ii. Simplifying Claims Process
a. People will remain essential to the claims process as automation evolves from process-centric automation to Artificial Intelligence; AI-enabled technologies are being deployed in almost every step of claims processes, including the digital intimation of claims, via APIs instant validation of basic facts directly with the Government data sources such as death registrar, collection of electronic health records through the hospital and the upcoming Ayushman Bharat Health Data-ecosystem, etc.
iii. Preventing Claims with IoT
Instantaneous data-sharing promoted by IoT and alternate data ecosystems also enables a fundamental shift in the relationship between insurers and customers, from the one in which ‘risk is transferred’ to a partnership with a shared interest to prevent losses.’ Connected devices such as wearable computers, health trackers, medical devices, etc., combined with third-party data will alert the customer to avoid the untimely death risk.
iv. Nurturing the Insurance Ecosystem for Better Claims Processing
Innovation in claims processing will accelerate as the insurance ecosystem, such as insurers, reinsurers, aggregators, distributors, AI service providers, alternate data providers, and many others. They will come together to co-create an ecosystem for data-sharing under a common regulatory and cybersecurity framework for cognitive engines to learn faster and expedite claim validation and settlement. A centralized distributed ledger-based repository of fraudulent claims will also help combat fraudsters.
India’s life insurance industry is on the cusp of a transformative era, powered by AI, a comprehensive data ecosystem, and advanced cognitive computing. These technologies are pivotal in bridging the country’s significant insurance gap, offering more personalized, accessible, and efficient insurance solutions to meet the diverse needs of its growing population.
