5 Prerequisites for Introducing AI into Life and Annuities

These are the steps that must be taken before artificial intelligence and machine learning can effectively be applied to life and annuities.

(Image source: Shutterstock.)

We’re hearing increasingly more about AI in insurance, with some segments of the industry being more advanced in their adoption than others. In some cases, it has enabled insurers to remove the person-to-person interaction altogether, making it fast and easy to purchase a policy from your preferred device in a matter of minutes, no printing lengthy forms or kitchen table meeting with an agent required.

But AI isn’t one-size-fits-all and must be gradually and thoughtfully implemented. For life insurance carriers and annuity providers, getting on board cannot be an “us too” move. It must be strategic and make sense within a specific context and with pre-defined goals in mind.

Beyond carrier readiness, we must first ask ourselves whether the industry as a whole is ready for AI and machine learning (ML). My take is that life and annuities insurance isn’t digital enough today. I’m not the first, and won’t be the last, to refer to this segment as a bit of a technology laggard. Our market doesn’t have all the digital infrastructure in place or the volume of digitized data necessary to meaningfully begin to weave AI and ML into operations.

These are the steps that must be taken before AI and ML can really shine in life and annuities.

Step 1: Ground your business in a digital-first strategy and mindset

Carriers won’t get the full benefit of AI, or potentially much benefit at all, if they’re not operating a large part of the business in the digital sphere. Think of it this way: AI isn’t magic. It’s simply working off data and building algorithms that make decisions informed by that data. The algorithms get smarter and more sophisticated over time as they’re exposed to more inputs.

If your back-end systems are offline and much of your information is still collected through paper-based forms, you can’t begin to pool the data (or enough of it) needed to apply AI and ML. Once your business strategy takes into account the need to be digital-first, you can start the process of digitizing systems and capturing the digitized information needed to apply AI and ML to.

Step 2: Digitize your data collection vehicles

This is where you ready your data onboarding and collection vehicles. Take it from consumers themselves; if it’s a long, slow process to get insured, you can bet they’ll abandon their theoretical shopping cart and move on.

Forrester’s report, “The State of Digital Insurance, 2021” said it well, “Customers have embraced digital touchpoints for a range of insurance activities. Where incumbents haven’t responded with better digital experiences, a host of insurance upstarts is ready to take over.”

An ideal data collection vehicle, for example, would be an eApp. Once you digitize rules and all of the applicant’s inputs are stored and analyzed digitally, eligibility decisions can eventually be made on the fly. This greatly speeds up the underwriting and purchasing process for you and them. A win-win.  All rules, knowledge and IP need to be digital in order to apply AI/ML. And where it makes sense to still have a human brain involved in making underwriting decisions (for example, to correct algorithmic bias that may be detected in the process), that’s where a person can add real value.

Step 3: Consolidate and integrate the data

Good decision making starts with good information. You can’t get good information if it’s fragmented and incomplete. By consolidating and integrating data into a single repository, data capture becomes seamless and decisions can be made more quickly.

A key challenge that stands before the life and annuities industry today is that it’s highly unstandardized. We need to get to a place where there’s more normalization across the industry around what and how data is captured, stored and used. This sort of utility model will ensure we’re all swimming in the same direction, finding ways to capitalize on AI and ML collectively and individually.

We’re seeing some progress toward this. Depository Trust and Clearing Corporation (DTCC), for example, is making inroads on cross-industry data standardization. In the meantime, a third party can help carriers with their own data strategy and standardization of data collection, using best practices from working with others.

Step 4: Ensure data quality

Once data is into a single repository, a data quality application should be run to ensure quality. The data cleansing process pinpoints and corrects any errors, such as removing duplicate inputs of information or identifying missing values. It then converts the data into standardized formats.

This is a critical step to ensuring consistency and completeness of your data before AI/ML algorithms can compare and analyze data sets, learning as they go.  Again, good decision making comes from good data.

Step 5: Put AI and ML to work

It is only once we’ve reached this point that we can truly refactor the way processes work by implementing AI and machine learning. This is where the life and annuities industry can transform in the ways we’ve seen in other segments of insurance do.

Take Lemonade for example. Not only is their claims submission process entirely digital and capable of being completed in mere seconds by the policyholder on virtually any device, their AI is able to approve or reject a claim momentarily. If the claim is approved, the payout is made within seconds, according to the company. If a claim is denied, that’s where a human rep gets involved to further investigate. This lets technology do the bulk of the work, but if a person is better suited to review and weigh in, that too is part of the process when called for.

There are many ways AI and ML can benefit insurers and policyholders, alike. Speeding up underwriting and calculating risk (perhaps eventually even in lieu of a medical exam) are obvious ones, for example. On the customer side, enhancing the user experience to be faster, more personalized and more engaging will be what sets one carrier apart from the next. Digitizing processes then applying AI and ML to speed up decision making will change the game when it comes to the timeline for policy holders to go from application submitted to coverage secured.

All of the players of today in life and annuities can consider themselves at risk of being edged out by the competition if they don’t embrace change. Utilizing AI and ML will help ensure carriers are future-proof and able to meet consumers where they are. But before we get to that point, the industry at-large has some work to do to lay the groundwork for enabling AI and machine learning to take hold.

Everybody LOVES Gelato (Data Science)

Brad Medd // Brad Medd is CTO at SE2, an insurance technology and service provider focused on the U.S. insurance and retirement industry. SE2 currently administers nearly 2 million active policies on behalf of its 25+ clients and has over $100 billion in assets under administration.

Leave a Comment