Assessment of the fraud detection technologies used by Kenyan insurance firms in detecting and preventing fraudulent insurance claims
Date
2023
Authors
Owuor, M. A.
Journal Title
Journal ISSN
Volume Title
Publisher
Strathmore University
Abstract
Insurance plays a vital role in the growth of economies, including Kenya's, as it serves as a mechanism to replace losses, which promotes sustainable economic development. Unfortunately, fraudsters have recognized the insurance industry's liquidity and susceptibility to fraud, leading to a significant influx of fraudulent activities. This has prompted insurance companies to employ advanced technology to detect and prevent fraudulent insurance claims. The goal of this research was to identify the fraud detection technologies that Kenyan insurance companies are employing to identify fraudulent claims, the firm characteristics that influence the selection of technologies, and the firms' perspectives on the efficiency of contemporary technologies in the identification of fraudulent insurance claims. The study employed a descriptive survey, and the data was collected through questionnaires. The population of the study was 56 licensed insurance firms in Kenya, as reported by IRA in 2021. It was established that the common fraud detection technologies include the Internet of Things (25.5 percent), machine learning (17.6 percent), artificial intelligence (17.6 percent), and blockchain (5.9 percent). Nonetheless, the insurance companies were using other forms of fraud detection (33.3 percent). The type of Insurance by-products is positively and statistically significant in influencing the adoption and implementation of fraud detection technologies. The combined ratio and size of the insurance company positively and statistically influence the adoption and implementation of fraud detection technologies by insurance companies. Board size and board independence are positively and statistically significant in influencing the adoption and implementation of fraud detection technologies. However, board gender diversity is not a significant predictor. The study concludes that the Internet of Things was the most popular technology used by Kenyan insurance firms to detect and prevent fraudulent insurance claims, followed, by machine learning, artificial intelligence, and blockchain. A conclusion is further made that the type of insurance by-products influences the adoption and implementation of fraud detection technologies. The size of the insurance company, profitability, board size, and board independence are significant company characteristics that influence the adoption and implementation of fraud detection technologies. Insurance detection technologies are useful and efficient in detecting fraudulent insurance claims. The study recommends the deployment of more than one technology informed by the different nature of fraud claims and dynamics associated with fraudsters and detection technologies. Before implementing any insurance fraud detection technology, the insurance company needs to consider the kind of insurance products it is dealing with. Moreover, profitability and company size define the ability of the insurance company to implement fraud-detecting technologies. There is a need for an effective board in the implementation of fraud detection technologies by insurance companies. With increasing insurance fraud claims, insurance companies ought to implement fraud detection technologies that suit the needs of the organization. The needs are parametrized by the usefulness and efficacy of the technologies in detecting fraudulent insurance claims.
Keywords: Fraudulent Insurance Claims, Insurance Fraud, Machine Learning, and Artificial Intelligence.
Description
Full - text thesis
Keywords
Citation
Owuor, M. A. (2023). Assessment of the fraud detection technologies used by Kenyan insurance firms in detecting and preventing fraudulent insurance claims [Strathmore University]. http://hdl.handle.net/11071/15391