December 10, 2018

Big insurers see opportunities, risks with AI, related tech

HBJ Photo | Matt Pilon
HBJ Photo | Matt Pilon
Insurance execs discussed evolving technology at the 2018 Insurance Market Summit. They included (from left) moderator Anand Rao and panelists Jonathan Bennett, Henna Karna and Santhosh Keshavan.
Matt Pilon

Spend some time at an insurance industry conference and you're likely to hear the term "artificial intelligence" or "AI" mentioned a lot.

AI itself isn't new — computers were beating humans at checkers around 1960 — but the technology has come a long way since then.

Insurers are already using AI-enabled chatbots to interact with their customers, and programs that can assess property damage through aerial or mobile photos to help speed up the claims process.

However, there are many more AI-enabled possibilities to explore, according to Anand Rao, global artificial intelligence leader for PricewaterhouseCoopers (PwC), who moderated a panel discussion on the topic at the recent annual Insurance Market Summit in Hartford.

"The technology is still nascent," Rao said. "But I think experimentation should be a part of the DNA [for insurers]."

Panelist Jonathan Bennett, chief financial officer and head of strategy for property & casualty and group benefits at The Hartford, said insurers have just begun to scratch the surface on how best to use AI and related technologies (like machine learning, deep learning and natural language processing) to make their operations more efficient and profitable and enhance customer relations.

Those tools can help process and analyze huge datasets, with the potential to find new insights and patterns, make more accurate predictions about customer behavior and claims, and even help identify new business models for insurers.

For Bennett, one area where AI holds promise is in analyzing the vast amounts of communications between his company and customers.

"Anything that can read and help us to understand what's been sent, how you can process that more effectively, make decisions before somebody needs to be involved to handle it, those are things that clearly add value and I think still have a lot of runway in front of them," Bennett said.

Panelist Santhosh Keshavan, chief information officer of Voya Financial, said his company is exploring the latest in voice and sentiment analysis, particularly as consumers increasingly use smart speakers like the Amazon Alexa to shop for products, including insurance policies.

Henna Karna, chief data officer at AXA XL, said the best use of AI is to enhance interactions and relationships with customers.

"If our data and tools allow us to do that, then it becomes a higher priority … ," Karna said.

While there are opportunities to use and help develop AI, there are plenty of risks too, said Rao, who showed a slide summarizing them. They range from cybersecurity and privacy concerns to the more futuristic (and scary) potential for AI weapons, drone swarms or systems that "go rogue."

One nearer-term concern for insurers is how their employees react to efforts to vet and implement new automation in the workplace.

Karna said she's seen employee skepticism, and thinks how management communicates its technology efforts is important.

"It's not about sunsetting jobs or specific workstreams, it's about 'how should we be operating in five years?' " she said. "We're not replacing any of it. We want to do it better."

Bennett said there can be cynicism when management has "irrational exuberance" about what AI and other technologies can do for the business, while not recognizing how much work needs to go into assessing those tools, some of which may not pan out.

"Not just because people are concerned about what it may mean to their job, but because they don't believe it can account for all the complexities that they face day to day," he said.

Most Popular on Facebook
Copyright 2017 New England Business Media