The use of public information creates privacy concerns.
Insurers can build greater trust in artificial intelligence (AI) by using customer-specific information and setting clear rules for how the technology handles data, analysts said.
“Using publicly available data for AI-based analytics involving customers can erode trust,” Olive Lam, senior vice-president at Lockton Companies LLC, told Insurance Asia.
Customers may worry that insurers are drawing conclusions about them from third-party sources, misreading sensitive information, or producing analysis that does not reflect their actual operations, Lam said.
The concerns matter as insurers expand AI use in areas such as customer service, claims and risk analysis. Only 38% of consumers considered AI helpful during insurance interactions, whilst 84% wanted companies to disclose when AI was being used, Smart Communications LLC said in a June report.
Lam said insurers could instead rely more heavily on information directly related to a customer, including historical claims and incidents, internal risk assessments, loss-control records, and site or facility characteristics.
Insurers could then combine those records with selected information from their own databases, including relevant loss experience, underwriting guidance, and comparable risks, she said.
That approach could make AI analysis more relevant and easier for customers to understand, whilst improving its reliability, Lam said.
The use of publicly available information could also create privacy concerns even when the data itself is not confidential, she said. Customers may question why information about them was collected and how it was used to reach a decision.
Oki Stefanus, a partner at EY-Parthenon Indonesia Strategy & Transactions, said insurers also need clear rules governing AI use.
“For insurance firms looking to integrate AI across their operations, establishing robust governance frameworks, transparency, and clear accountability is essential to earn the trust not only of customers but also of other stakeholders,” he told Insurance Asia.
Stefanus said insurers should address data privacy, third-party data use, and regulatory requirements when deploying AI. They should also regularly review security measures as AI systems change, he said.
Clear explanations of how AI is used, how decisions are made, and how risks are managed will also be important, Stefanus said.
The Smart Communications report found that 29% of consumers had stopped doing business with an insurer because of data privacy or transparency concerns. About 63% of consumers globally said they would switch insurers if communications failed to meet expectations, rising to 67% in Singapore.
Lam said insurers that use information closely tied to customers’ actual risks could give AI a clearer basis for analysis instead of relying heavily on broad third-party data.
Stefanus said responsible use of AI should ultimately improve results for both customers and insurers, but accountability and clear communication should remain part of its deployment.