Kathmandu. Insurance companies are increasing the use of artificial intelligence (AI) in their operations. However, experts believe that this technology should be used only to help in the claims process. They believe that the use of AI is not appropriate to determine whether to pay a claim or not.
According to Ross Sinclair, chief executive of EIP Ltd., there is a big difference between using AI to gather information, detect errors, or process simple queries, and allowing AI to make final payment decisions. He suggested that insurance companies should use their own rules to automate the claims processing steps. However, he said that final claims decisions should be completely human and company control.
A recent survey by S&P Global Ratings of 121 insurance and reinsurance companies around the world shows that 81 percent of companies are prioritizing automation and 71 percent are already using AI to improve their internal support functions. Half of these companies are focusing on customer service. While 39 percent are using this technology for risk management tasks such as fraud detection and pricing.
Despite this, 83 percent of the companies still do not fully adopt AI. They are still in the early or mid-stages. Only 8 percent of companies have fully implemented AI projects in their entire organization.
Insurance companies expect efficiency improvements of up to 7 per cent and revenue growth of up to 5 per cent by 2028. The number of companies that expect cost savings of more than 3 per cent with AI is expected to increase from 16 per cent in 2025 to 80 per cent by 2028.
According to S&P analyst Andreas Lindberg, operating benefits in the areas of claims, underwriting and customer service are more pronounced than financial gains. So far, this technology has not significantly affected the profitability or credit valuation of insurance companies.
More than three-quarters of the companies surveyed cited data privacy, security, and regulatory compliance as the main barriers to widespread adoption of AI. About 62 percent of companies list AI models to monitor their use and associated risks. As the industry expands on these tools, one of the biggest challenges is ensuring that AI increases efficiency and accuracy. However, individual claims do not control decision-making. – Agency
