Where does AI see itself in the next five years?

Anthropic, Allianz partnership proceeds with underwriting automation

Anthropic and Allianz point to careful automation of underwriting as a means to advance responsible AI, in executives remarks at the Insurtech Insights conference in New York on June 3. In January, Anthropic and Allianz announced a partnership to implement AI for the insurer's operations. 

Harness design, which is part of how Anthropic's Claude AI builds applications, allows underwriters to concentrate on high value tasks, said Mike Ram, head of insurance at Anthropic.

AI has become a bigger risk for insurers than business interruption, climate change or natural catastrophes, said Christian Freytag, head of group technology and data, and group CTO at Allianz. In the next two to three years, defending against cyber attack capabilities powered by quantum computing combined with AI, is an area where insurers will need to "double down," he said.

Being able to trust AI technology is important, Freytag added, and insurers need to identify the intelligence that AI systems are accumulating, because rules of insurance operations always run behind what the technology can do.

New York Life CIO expects 'substantial' AI impact in five years

New York Life expects AI to have a "substantial" impact on operations in the next five years, said Deepa Soni, executive vice president and CIO, in remarks at Insurtech Insights. When thinking about where friction is with end users or advisors, and the time they should be spending with customers versus dealing with data – all of that can be simplified as we think about elevating the experience, Soni said. AI can track where customers are in a transaction's lifecycle and flag customers who should be called by a person, she added.

Proceed methodically and seek value, TIAA and Tokio Marine leaders say

Insurance company technology leaders will have to figure out their own value and proceed methodically when working with AI, according to tech executives from Tokio Marine and TIAA who spoke on a panel together at Insurtech Insights.

"If you're on the value chain for delivering technology, as that delivery becomes more agent-ified, where do you see your value?" said Sastry Durvasula, chief operating, information and digital officer, TIAA. "How do you advance from that point, help the business transform and actually create AI data opportunities from a digital business models perspective? It's a great opportunity, so you can be on the front foot or the back foot."

Being in the business of lowering risk, insurers must not "suddenly become very risky in our own behavior" when using AI, said Robert Pick, EVP & CIO, Tokio Marine North America Services.

"We have to balance the opportunities that we have today, and where we're going. There's an appropriate degree of measuredness, and also allowing governance, safety and observation to catch up a little bit to the capabilities," Pick said. "Making sure that we're doing things methodically, in a way that progressively moves us in a direction, while businesses that can move more quickly, please do. Those that need a little bit more time, as long as you're making use of AI in the highest and best way you can on managing risk and cost – phenomenal."