Insurers have been debating ways to leverage AI in their insurance operations for quite some time, and these conversations have become even more fraught with the advent of Gen AI. The inability to make these decisions quickly is understandable when we consider that insurers are often quite risk averse, face significant regulation, deal with fragments of legacy IT systems, operate at massive scale, face meaningful brand risk, have many competing priorities, and very complex, interdependent systems and operations.
Common questions are
-
“Can we trust AI?”
-
“What happens if….?”
-
“What should we do first?”
-
“How will this affect customer experience?”
-
“What are the benefits and risks of being a first mover?”
-
“Do we have the technical know-how to do this on our own?”
So how can Insurers approach AI?
Insurers are certainly no strangers to technology. However, artificial intelligence is different. It’s often seen by regulators, customers, and even employees as a “black box” due to the challenges associated with explaining how it makes “decisions,” which is exacerbated by the inexplicability of many of Gen AI’s hallucinations. Insurers are understandably unwilling to tolerate the wide-ranging risks associated with these uncertainties.
That said, customers have come to expect more from their insurers. They expect real-time and fully accurate quotes. They want to be able to easily see, in real time, how adding a coverage, decreasing a deductible or policy limit, adding a driver, or getting a home alarm or water leak detection system would impact their premium. They want to be able to get accurate and understandable answers to questions quickly, easily, and via any channel which they choose to use at that moment. They also expect a meaningful response to their first notice of loss (FNOL) now, not in a couple of days.
Leveraging AI is no longer optional, and it isn’t simply for the sake of using the latest technology. It’s to meet customers’ service expectations, identify potential fraud, better match price-to-risk, manage operational expenses, remain relevant, attract and retain policyholders and employees, and compete effectively and efficiently in a highly dynamic marketplace.
Now is the time to dive in. Start slowly, but not at a single point. Consider starting with some internal-facing use cases to reduce risks and learn. This will also help you determine whether you should ‘build or buy’ to leverage commercial components versus a more unique in-house approach for certain other processes. Then test various external-facing use cases across a range of access points, learning and adjusting in real time. This can further reduce risk, increase learning, and get you to a safer, broader deployment more quickly than a piecemeal approach.
Ready to dig deeper? Here are 3 articles you might be interested in:
- CB Insights’ Insurance AI Readiness Index scores 49 insurers on Innovation and Execution. Find your company, as well as your top competitors, and think of examples of what may be driving these scores, and how you may be able to move the needle to match your company’s aspirations – https://www.cbinsights.com/research/ai-readiness-index-insurance
- The National Association of Insurance (NAIC) has a great deal of useful information about utilizing AI in insurance – https://content.naic.org/insurance-topics/artificial-intelligence
- Business News Daily published a great article “How Machine Learning is Transforming Underwriting” – https://www.businessnewsdaily.com/10203-artificial-intelligence-insurance-industry.html
Note: this article was originally posted by Bill Cecil on LinkedIn