What do insurers think about AI?

March 26, 2024 6:42 AM

AI radio testbed for the spectrum collaboration challenge from the U.S. Defense Advanced Research Projects Agency displayed — The artificial intelligence (AI) radio testbed for the spectrum collaboration challenge from the U.S. Defense Advanced Research Projects Agency (DARPA) is displayed during the Mobile World Congress Americas event in Los Angeles, California, U.S., on Tuesday, Oct. 22, 2019. The conference features prominent executives representing mobile operators, device manufacturers, technology providers, vendors and content owners from across the world.

Recent Arizent research shows insurance organizations may be more familiar with artificial intelligence than other industries. But while there are use cases for the technology and organizations are deploying them, some insurers are concerned AI will cause more customer experience woes.

The research was conducted online by Arizent in January 2023 among 1,065 professionals across several verticals including banking, wealth management and insurance. There were 77 respondents from the insurance industry. The research was sponsored by Salesforce.

The top AI tools used in insurance include ChatGPT, 74%; Bing AI, 48%; and Bard, 39%. Despite some concerns, 43% of insurance respondents said that AI is evolving at the right place. While 34% said it was evolving a little too fast.

Not all industries have implemented chatbots for employees, or copilots, but insurers seem to be ahead using the technology most frequently to capture and transcribe meeting notes, track action items, manage day-to-day tasks and for hiring and interviewing.

The survey asked respondents several questions related to AI. Digital Insurance gathered some of those comments.

What are the long-term ramifications or risks you are watching or are most concerned about with the potential introduction of generative AI in your industry or profession?

1. "I think the biggest risk is that it gets over-implemented and we lose the personal touch with our end-users. One of the things people love about us is our personalized levels of support. We do provide a tech platform with AI features and automation but there is always a person to speak to if you get stuck. In a world where AI is everywhere and people start interacting with machines rather than people, I think the ability to speak to someone and make a human connection could actually become an important difference."

2. "'AI can generate a lot more content than humans, but that content can be generic and low-quality. AI can also hallucinate and make up facts, which has potential to misinform readers. Perhaps I am biased as someone who has mostly made my living writing and creating content, but I think replacing human content creators with AI would be a huge mistake at this point."

3. "In insurance Gen AI poses potential risks related to the generation of deepfake content, which could be used for fraudulent claims or misinformation. Ensuring the security and authenticity of data becomes crucial, as malicious actors might exploit generative AI capabilities to create convincing but false information."

Regardless of your company’s plans regarding the adoption of generative AI, how is your company’s leadership team positioning the use of generative AI tools/systems within your organization to its employees?

1. "Leadership has encouraged the use of AI and even held a contest for employees to submit ideas for using it within their respective areas and awarded prizes to the winners."

2. "Cautiously optimistic, AI has huge potential to drive innovation and competitive advantage if done the right way with data protection, security and maintaining a commitment to ethical practices and risk management."

3. "Every department has a mandate to find ways to generate efficiencies from AI. We have a hiring freeze at the moment so if I want my department to do more I have to use artificial intelligence to bridge the gap. Our CEO is very in touch with developments in the fast moving areas of AI development and will be in touch to flag tools he thinks are useful and help specific departments or staff members."

What steps, if any, is your organization taking to make sure the generative-AI powered chatbots you are building for customers are aligned with customer needs and delivering the right experience? Please be specific and list all that apply.

1. "We are looking at ways to build out the knowledge base that the chatbot is drawing on so that it can provide more specific and useful information. There are tools out there that can do this where the knowledge base is private rather than public. We are also looking at the right prompts to create the personality of the chatbot - one requirement being the need for simple direct language."

2. "We are looking for tools that will not expose other customer's info and will not provide incorrect info. However, many customers are old, and they may not like it, and prefer more human interactions. We may have to wait several years before we achieve appropriate customer turnover and have a customer mix who will better appreciate using customer-facing AI tools."

3. "Our chatbots on the policyholder website are not as advanced as other companies. However, I think there is a focus on how they can be more dynamic for the customer, in hopes that they can resolve customer issues and answer questions to help build brand loyalty and retention."