The insurance industry has been warned over the limitations of artificial intelligence as it continues its digital transformation.

AI has generated great enthusiasm among industry executives and IT service providers, but it is important to understand it operates by predicting the data or references it should utilise next without possessing actual knowledge, according to global researcher Information Services Group.

“At its essence, AI is a predictive model that leverages substantial amounts of data,” the group says in a new report. “The ability to cultivate and apply data with an appropriate level of granularity will be a critical factor in determining AI readiness among insurers.”

ISG notes AI is designed to perform only within its programmed parameters.

“Thus, one should not rely on AI to deliver intelligent responses or solutions that extend beyond its pre-defined capabilities. Furthermore, a lack of access to first-party data and ongoing concerns regarding data privacy will continue to pose challenges for insurers, as most AI systems were not originally developed with security in mind.”

Insurance companies in Australia and New Zealand are embracing new technologies amid declining margins, with inflation and extreme weather affecting claims, according to the report.

The industry also faces pressure from changing consumer expectations and competition from start-ups that is forcing incumbents to innovate in product and customer experience.

“Insurance companies in Australia and New Zealand need to move beyond legacy systems and cultures to remain competitive,” ISG partner and head of Asia-Pacific Michael Gale said. “Service providers are helping them make essential leaps in technology.”

ISG says AI and generative AI are playing a growing role in automated claims processing, better customer interactions and predictive analytics for risk management. Property and casualty insurers are using data analytics to improve risk assessment and customer segmentation.