More than half of insurers use machine learning in analytics

Register now

More than half of today’s insurance companies use machine learning for predictive analytics, according to a new report by Earnix, an analytics software provider for the financial services industry.

Roughly 200 insurers were surveyed as part of Earnix’s global “Machine Learning: Growing, Promising, Challenging” study, and they were prompted to select all business areas applicable to them. In total, 70% deployed the technology for risk modeling, the study found. Other uses include the creation of demand models, which forecast market appetite for new products, at 45%. Fraud detection was the third-most popular application, used by 36%.

“Sometimes there are passing technology fads, but [we] believe that machine learning is here to stay,” the report says. “If you are one of the 46% of companies that currently has not investigated the power of machine learning, the time to start is now.”

Industry consensus is machine learning will bring significant change to insurance over the next five years, with 71% of companies believing investments in the technology will increase, Earnix says. Nearly 60% of early adopters claim machine learning has already made their analytics models far more accurate, leading to better risk assessments and underwriting decisions.

However, the emerging tech is not without its obstacles. Almost half of carriers cite a lack of knowledge or skill as the main impediment to adoption. North of 80% of insurance executives also confess their organization is relatively inexperienced when it comes to machine learning. Additional challenges include a shortage of deployment tools and trust in methodology, according to the report.

“Fortunately, these barriers are just short-term. Like with any new technology, it takes time to build both the internal competencies to execute and the business buy in,” Earnix says. “Machine learning is positioned to have a significant impact on the way insurers do business.”

For reprint and licensing requests for this article, click here.