Legal & General's transformation invasion

Legal and General America headquarters
Legal and General America's corporate headquarters, 3275 Bennett Creek Avenue, Frederick, Maryland.

Raju Seetharaman, chief technology officer of life insurer Legal & General America (LGA), began his role in 2018 after 18 years with the U.K.-based global financial services company's London office. There, he built its digital platform and also served as CTO. Seetharaman and Mark Holweger, CEO, who also came from the parent company, are applying technology and digital transformation to grow its U.S. business. One area where LGA has already made a tech-based transformation is in reducing the time needed to write policies for life insurance customers. Digital Insurance heard Seetharaman's perspective on how data and AI supports digital transformation for life insurance.

This interview has been edited for clarity.

How do the needs of the U.S. and U.K. markets differ, and how does that affect digital transformation?

Raju Seetharaman of Legal & General America
Raju Seetharaman, chief technology officer of Legal & General America.
I like the ability in the U.S. to have access to data. In the U.K., we have GDPR data privacy rules. The U.S. is open to sharing data and using data more. That is a significant advantage. The first time we did automated underwriting in the U.K. was in 2002. It took a long time to get to 80% straight-through processing.

We launched our automated underwriting in 2019. The progress is because of the access to data and good quality medical evidence data. It's evolving continuously. There are a lot of innovators in that space. Data is a big differentiator – how data is available and accessible, and how that can power innovation – is a key difference.

What data has become available or have you started collecting?

Today we get real-time prescription history data, which is not available in the U.K. We use six different sources of data and medical information. The idea here is to help the customer get the policy on the same day or the same session. 

Let's say I'm applying and someone asks, 'Do you have high blood pressure? I say no. 'Does your family have it?' Yes, they have a history of high blood pressure. 'Do you know your latest reading?' 'I don't know. I can't remember.' 

What if you can give [a blood pressure reading] to them, sourcing it from an electronic health record? Then the person says, 'Yes, I see that. I agree.' Then it gets pushed into that, you get an instant issue. If I didn't know and said I didn't know, the underwriter may say they want to know, go and take your value and come back. This accelerates the good customer experience we can provide by having access to real-time data, or sometimes even near-time data.

Are there differences in how you reach personnel, including brokers and underwriters, with transformation?

The challenge we faced initially was agents concerned that they are being disintermediated. If you look at insurance fintech, they all started by saying, we are going to directly reach out to the customer. The initial challenge is building a customer app where a customer can complete the life insurance application without telephoning them and a stranger asking very personal medical questions. Once you decide on the carrier and the price, then they're asking detailed questions [like] have you smoked marijuana, do you have an STD, or do you have cancer? Your cancer might be in remission and you don't want to talk about it.

It should be an intermediated journey. Agents are the experts. It took a long time to convince them, digital doesn't mean you're disintermediated. Digital means you don't need to touch it. You get the money without touching. And that makes you a profit. 

That is a challenge our company had to overcome. I told the board, we made the investment and built the platform. They see the demo. It looks good and the data looks good. They ask why adoption is slow. I said I had a first mover disadvantage – counter to the typical, when first movers have an advantage. If you are an agent dealing with eight carriers that are analog, only using phone or paper, and suddenly one becomes digital, you have to adjust your business processes to deal with that one that's digital and seven others that are still analog. That is not easy for the industry, so we need to coach them to get there. Luckily, other carriers are starting to pick up, which will help.

AI’s potential for all industries is often promoted. In the insurance industry, is AI just another tool in the toolbox for customer service, underwriting and other tasks? Or is it an important advance for innovation and transformation?

It's a bit unfair to look at AI as one monolith. It has multiple nuanced layers. Gartner talks about everyday AI and game-changing AI. LGA plays in both spaces. In LIMRA's quarterly reports on retail individual life insurance sales, we went from eighth among carriers when I started [in 2018] to third among carriers as of the first quarter of 2024. That is because we're using digital transformation to access customers. Our ability to work with our customer, listen to the customer and deliver improvements, keeps us ahead of the marketplace, but we are not standing still. We need to move forward. 

Data and AI are crucial parts of moving forward. Data plus AI is very, very powerful, because everybody can have a large language model, for example, and it's accessible. It's subsidized as well. Data is getting the differentiator in that if you use data and prompt engineering appropriately, you can create your own intellectual property, which is hard to replicate. That's the space we are operating under.

Human beings are good at thinking. Human beings are doing. Some days I'm only thinking, some days I'm doing and thinking. Some days I'm only doing. Today is all doing. There's no time to think. AI is going to help us to move away from the doing part to the thinking part, which is innovation -- our ability to do things that will be hard to replicate. 

Productivity is one of the core concerns at a macroeconomic level. Productivity improvements and having operational leverage for growth are key areas where AI will be super helpful for companies. We have gone from approximately $100 million to $200 million in annualized premium equivalents, doubling our business, But we have not doubled our workforce to achieve that growth. We are able to use automation and digital transformation to power our growth. If we can add AI to that mix, we can grow to $300 million and our talented workforce can evolve towards more thinking than doing. 

Game changing takes complex medical information and divides it into consumable, small bite-size information to an underwriter so they can overlay their expertise to make the right underwriting decision for a customer. Today, my underwriter has to read 200 pages, scan it, and understand what is disclosed or not. AI can do all of that for them and show them that this is a focus area. People's time on underwriting can shrink really, really small.

How will AI improve thinking, an area where it has had problems? It’s good at collecting information and finding things. What areas in insurance need more thought or would benefit from applying AI?

The first step is to look at AI, and it's in the infancy stages. It's like a kid at this stage in that it needs guidance. It needs management. If you go back 100 years, the automobile was a bit of a wild thing to control and drive. AI sometimes does not do what you want if you are not giving it the right instruction. That's where the problem is, in the layer where that innovation happens, about how we are grounding the data. There are many techniques, which are in the leading edge of how to get the maximum capability out of the AI. We are in early stages. 

So going back to the point, that's a foundational thing. It's an evolving journey. You have two opportunities. You can innovate with it, or you can be the first follower. Wait for it to mature and then be the first adopter. We are deciding to be a fast follower or follow the trend. We'll be the leading edge of it, because we already have a digital, innovative platform allowing us to grow. So we're running it in parallel and using it to innovate. 

Going back to the life insurance value chain, life insurance buying uses self-declared information. We are validating that information by accessing data from resources. That's a turnkey two-part process. We want to merge both together. AI will allow us to do that, by getting up-to-data data and adapting questionnaires to ask only targeted questions. That's the kind of innovation we are looking for, where we can shrink the application experience to a smaller experience so the applicant doesn't need to answer 55 questions.

What else are you seeing in technology for life insurance that has you thinking about its applications for LGA?

Large language models have created the capability to interact at a human-like level. The next step, happening now, is, how do I change voice to text and text to LLM. That is where I see more evolution, in life insurance in particular. Why do I need to type something? Why can't I say something? 

Accents are a problem. Tools struggle with accents. The AI element can interpret different accents more appropriately, allowing us to engage more realistically, and I see that as a big area of development. The other area of development is how we leverage data. Data will continue to be the core of innovation. How do you monetize your own data? Enterprise data is something we continue to talk about and find ways to leverage, because that's what is going to give us the market advantage and competitive advantage.