As insurance continues to evolve, the link between analytics, AI and organizational success has never been more apparent.
Marvin Pestcoe serves on the boards of Hamilton Insurance and Catalina Insurance. He is a fellow of the Casualty Actuarial Society. In short, he understands the connection between analytics, AI and insurance organizational success.
In the wake of the excitement around generative AI (GAI), ChatGPT and large language models (LLM), Pestcoe positioned these new tools against his past experiences of how advanced insurers have implemented analytics. His thoughts provide a roadmap for insurance leadership. Three of his main points are summarized below.
1. The generative AI opportunity
Pestcoe confirms that GAI is a top issue for insurance company boards of directors. He hopes this attention translates into a renewed focus on analytics. Since GAI and LLMs aren’t new, many underutilized tools and techniques are already in the market that are proven, tested and effective. These need to be better mined for their full value. Pestcoe is hopeful that the attention around new AI tools will help raise the discussion of analytics in general.
Related: Adjusting to disruption and uncertainty in insurance has become the norm. Learn ways that your insurance organization can stay resilient amid these changes.
2. Defer to technical experts at your own risk
One of the inhibitors for analytics in insurance is that, for many insurers, analytics has been “shunted off into the back room.” Pestcoe does not believe that analytics has received as much focus at the board level as warranted for something critical to an insurer’s success. His surprising reason is that insurers became entrenched in traditional techniques and tools because they were early adopters of analytics.
As new approaches such as machine learning became available, they were seen as “nothing new here.” Pestcoe observes that he has seen boards give too much deference to technical experts and encourages board members to challenge technicians to adopt new approaches.
3. Two changes boards will expect
The importance of analytics across the insurance enterprise will cause boards to change their approach to information technology issues, Pestcoe predicts. He sees two major adjustments:
- First, boards will expect more thorough briefings from management on IT issues. They will challenge management to present the business implications of information technology issues clearly enough so that boards can weigh in with informed decisions.
- Second, boards will expect senior managers – CEOs, COOs, CROs, etc. – to own information technology issues and not hand them off to back-office technical experts.
Pestcoe closes his remarks by highlighting three main areas of risk in analytics – bias, control and trust. If data is biased, analytic results will be biased and improper use will occur in the absence of control. Without an emphasis on trust, some insurers will lose their customers’ confidence. The new GAI and LLM tools further heighten these existing issues. Pestcoe expects that boards, senior management, and insurance solution providers will collaborate closely to address the issues and mitigate risks.
