A tech solution to the fraud conundrum
04 Apr 2022
A geographically diverse and complex country like India makes the business of insurance interesting. Its cultural multiplicity, disparity in standards of living, and the vast variance in opportunities across geographies offers layers of insights into how each customer perceives financial risks. It is also what makes the on-ground reality of carrying on the insurance business fairly challenging.
For decades now, fraud has been among the major challenges facing the sector. Traditionally, frauds have occurred more in rural and semi-urban areas, where insurers lack a robust infrastructure to inspect suspect cases. These frauds, which may have been carried out individually, have over a period of time evolved into large undertakings of organized syndicates.
The modus operandi of such syndicates ranges from using the details of terminally ill individuals to benefit from their inevitable deaths, to filing false road accident claims of people who died of natural causes. These syndicates function with the sophistication and efficiency of an organized company, with skilled men and women making concerted efforts to defraud insurers and genuine customers. Traditionally, several fraud cases from a single region often led to insurers taking multiple steps to stem the flow of frauds, which may at times lead to loss of access for even genuine customers.
A sea of data
In the pre-digitization era, despite heavy oversight, it was nearly impossible to effectively monitor, investigate and identify such fraud incidences given a large amount of information at play. This ever-growing pool of data came from multiple source points and companies relied on human intervention to make sense of this data.
Matters worsened because this information was received and processed in silos, increasing the likelihood of errors. The biggest impact, quite naturally, was the financial one. In 2019, life insurers are likely to have suffered losses worth Rs. 45,000 crores due to fraudulent claims alone. A report by Lancers Network and Insurance Institute of India estimates that the past two years have seen a further increase in such incidences, with over 50% of companies estimating a 30% increase in fraud cases.
However, the quickened pace of digitization during the COVID-19 pandemic has been a game-changer. New-age solutions like AI, Big Data, and Machine Learning have offered an effective and faster method of analyzing information and detecting fraud early on in the customer lifecycle. Insurers have been able to not only understand customer behaviour better, but also have gained a competitive advantage by adopting these digital solutions.
Impact of Digital
The biggest boon provided by digitization has been the seamless integration of information. The industry has already made significant strides and in the last 2 years especially have redefined these critical processes.
Insurers are now able to build an enterprise-wide perspective on anti-fraud efforts. Analytics has helped in linking associated data within the organization and enabled insurers to monitor customers throughout their life journey and create identifiers to separate fraudulent behaviour from a genuine one.
Fraud, typically, occurs at multiple source points – claims, surrender, premium, application, employee-related or third-party. Technologies like Machine learning and analytics build a predictive value by combining data from each source point and suggest the propensity of occurrence of fraud in each individual case based on these identifiers
These analytical solutions are helping reduce the overall cost of fraud detection by proactively catching it earlier in the lifecycle. This has a cascading effect on the issuance of policies, processing of legitimate claims, etc. In India, many life insurers are now performing analytics with Big Data to flag or validate claims. This automation process offers a dual benefit – not only does it help in delivering real-time results, but also doesn’t compromise on thoroughness.
The road ahead
In future, AI will provide a competitive advantage to insurers, as it will help gain an understanding of customers on an individual level. This will prevent the exclusion of genuine customers from gaining access to insurance. With these new technologies, insurers can proactively separate genuine customers from fraudulent ones, with a higher level of reliability, thereby creating more democratic access to life insurance. It will truly create a seamless customer experience, as risk profiles will become more holistic, and companies will be able to deduct actionable insights from the data already at their disposal.
These new technologies, however, haven’t fully eliminated the need for human intervention. In fact, it has led to a newer, more specialized function within companies which dedicatedly works on creating solutions that can flag fraudulent incidents early in the customer lifecycle. This team also looks at improving existing models to determine what is normal customer behaviour and what is potentially fraudulent. These human insights are helping make the existing models more robust over time.
While data is at the core of this technology revolution, insurance will inherently remain an industry that relies on gut feeling and human insight. A proper mix of automation and manual intervention can bring fraud detection to a new level and an analytics backbone can assure the highest level of objectivity.