Intelligent automation drives operational insurance improvements

Insurers looking toward digital innovations to drive operational improvements in their processes would do well to consider the benefits of intelligent automation (IA). For example, the recent partnership between SilverBridge and Astute Financial Services Exchange which delivers a fully automated intelligent claims processing solution that provides maximum risk oversight while enabling claims to be processed in real-time. It centres on leveraging innovative technology to help reduce the cost of each claim thanks to better ways of detecting fraud.

In fact, managing fraud remains one of the most significant challenges that insurers are faced with today. In South Africa, the latest figures show that commercial crime, of which fraud is a contributor, is the fastest-growing segment in the country, increasing by more than 14% year-on-year. Given the difficult economic conditions resulting from the lockdown, the expectation is that instances of fraud will likely increase this year. Meanwhile in the US, the total cost of insurance fraud, not including health insurance, is estimated to be more than $40 billion per year. This is such a universal industry problem, that the global insurance fraud detection market size is forecast to reach more than $7.9 billion by 2024, growing 26% annually from 2019 to 2024. Hardly surprising then that digital advances to curb this have been welcomed.

Combating fraud

IA combines robotic process automation (RPA) and artificial intelligence (AI) technologies to facilitate rapid end-to-end business process automation and accelerate digital transformation. For its part, AI has become sufficiently proficient at processing unstructured data using complex algorithms. The amount of historical data available at an insurer means that they can improve processes, enhance the customer experience, monetise data, and detect fraud in more innovative ways.

By using the advanced data and natural language processing (NLP) capabilities of AI to mine data from claim forms, fraudulent claims and patterns of fraud can be flagged. Furthermore, AI can identify patterns through the syntax of the descriptions of incidents from several different claimants to detect organised fraud. When combined with RPA, insurers can benefit from the automation and simplification of complex business processes to help virtualise human decision-making.

By continually analysing internal and external data in real-time, IA creates an environment that continually learns and evolves from the data available to it. This can cross-reference and analyse any number of internal and external databases, detecting potential patterns of abuse. Linked with the replication and automation of decision making, IA creates a scenario of increased efficiencies, consistency, and accuracy.

A famous example of this is the sudden increase in broken or stolen TV claims that appeared simultaneously with the introduction of new technology and prior to major sporting events, in this instance the 2016 FIFA World Cup.

In another example, one of the largest insurers in Turkey used to employ a team of 50 people to manually check each claim for fraud based on loose rules and the team’s personal experiences. But considering that there could be up to 30 000 claims a month, this was becoming increasingly impractical. Using IA, the insurer was able to save almost $6 million in fraud detection and prevention costs and significantly improve the customer experience in the process.

Driving down claims costs

As far back as 2017, IA has been found to deliver cost savings of up to 75% with the payback ranging from several months to a few years. With insurers continually examining ways to optimise cost efficiencies, the savings available from IA could mean being able to invest in growing other business opportunities.

Take for instance how NLP can rapidly mine information contained in handwritten notes that go with either claim or insurance-related forms. In healthcare, these could be relevant to standard procedures and diagnostic terms, including abbreviations and diagnostic terms used in writing test results. Once this information is extracted, IA can automatically identify what is missing, classify it, and route it for the best action to be taken.

On the commercial insurance side, the ability to search through enormous quantities of data to detect a variety of fraud methodologies can result in decreased premiums as any objectionable claims are identified before being processed and paid out.

IA will continue to evolve as technology improves, making it even more accurate. By improving its ability to deliver better analysis using faster data, processing capabilities will assist the insurance industry in positioning itself for growth in more exciting ways while being less concerned about managing the challenge of fraud.

Comments: