Big data, analytics comes before artificial intelligence: panel
Insurers are keen to adapt artificial intelligence in the enterprise, but first they must build up a layer of data and analytics excellence.
That was the sentiment in a panel discussion on artificial intelligence held at the Global Insurance Symposium in Des Moines, Iowa Wednesday.
"People are talking about not just big data, but fast data, how we need to get that data and make decisions," says Anand Rao, partner and innovation lead for PwC. "We need to build trust in these AI systems -- they need to be transparent in how they interact."
Chris Augery, CEO of the insurtech startup Drive Spotter, says that when his company builds models based on the driving data it collects, it anonymizes it. Specific data are owned by the customer, he explains.
"Raw data is owned by customers and perpetually licensed to us. That allows us to sell redacted versions of that data to insurance companies," he says.
The proliferation of data is good for insurers in many ways, as it helps them automate some processes and make a better customer experience, added Louise Billmeyer, CIO for Principal Financial Group. But insurers must remember their ethical commitment to customers as well, she adds.
"In the underwriting process we're using information every day to make a decision," Billmeyer says. "When you think about just gathering that data, we're looking at biometrics, electronic medical records -- that still has to be at the consent of the customer."