? Searching for a way to cater to younger adults, one insurer is offering a new product that uses details gleaned from photos to accelerate its life insurance quote process.

? Within several years, facial analytics may be used to predict the onset of a number of chronic diseases, potentially creating a powerful tool for insurers to use when engaging with customers.

After reading a LIMRA analysis that showed that many millennials would be willing to wear biometric trackers when interacting with insurance companies, Legal & General America Inc. Executive Vice President, Business Strategy and Innovation Jim Galli took action to capitalize on new technological trends.

Following roughly a yearlong process, Legal & General launched a new website with Lapetus Solutions Inc. in late July that provides a customer with a life insurance quote based on details garnered from a "selfie" photo. The website, SelfieQuote.com, uses facial analytics technology to estimate an individual's age, gender and body mass index. Consumers are then asked to complete a quick survey that helps to generate an insurance quote.

In an interview with S&P Global Market Intelligence, Galli and Lapetus Solutions Chief Marketing Officer Janet Anderson discussed the ramifications of the advances in facial analytics and insurtech on the life insurance industry. The following is an edited transcript of that conversation.

S&P Global Market Intelligence: The process of incorporating facial analytics into the life insurance quoting system is virtually unprecedented; how did the idea come about?

Jim Galli: It's fascinating technology. It's fun, it's engaging. The value is in you telling us your rough age, then we'll be able to get you a quote based on the simple assumption of $100,000 for a 20-year term. You can play with the levers and say, "Well I might need a little more, what about a 30-year or a 10-year?" The goal is to see if it engages more customers and educates them. You could be pleasantly surprised if you run some of these quotes. If you're 30, 35, or 40, $100,000 is not that much money.

How do you plan on accounting for certain disparities between the age predicted by the platform and a consumer's actual age when vast differences arise?

Galli: The more volume we get, those extremes will be minimized. A couple people in the building here came out older, a couple people came out significantly younger and were happy, of course.

We do give tips, if you take a photo that doesn't work, ensure your face is well lit, remove your glasses, push your hair away from your face, look straight into the camera and keep it at an arms' length. We never said it was like a carnival where we will guess your height, weight and age 100% [correctly]. There are pluses and minuses; there will be a few outliers at the end.

A longitudinal study on facial analytics showed that the accuracy rate for identifying smokers through the technology is about 85%, while the rate for correctly estimating a consumer's BMI is around 80%. Do the findings indicate that Lapetus Solutions' platform is reasonably accurate in helping determine a quote?

Janet Anderson, chief marketing officer at Lapetus Solutions Source: Legal & General America

Janet Anderson: Each part of our process is built upon specific algorithms. We have nine different scientists that are working on the algorithms to provide a degree of confidence to insurance carriers that what we're able to come out with fits within a range of accuracy. Again, with machine learning and [artificial intelligence] techniques, the more data you have the closer you get to 100%. You will never get to 100%, but we want to make sure that we'll be able to provide a high degree of confidence for carriers to help them make their decisions possibly cheaper, faster and [provide] a better experience for the consumer.

Lapetus Solutions co-Founder Jay Olshansky has mentioned that facial analytics technology may eventually be used to predict a wide range of medical conditions. How quickly can this be accomplished?

Anderson: That is definitely on our roadmap. It's not something that happens overnight, it takes a lot of what we call ground troop data. We take a series of images. We take still images, we take different angles of the face and we also take some video. With this data, we'll have health data, as well as images. This will be the foundation of what we're able to utilize to build out algorithms that can detect chronic diseases such as heart disease or diabetes or perhaps some sort of dementia. Ideally, we want this to be done within two years, but it depends on the data. We can't rush the studies because something like this can really help the consumer get insurance a lot faster.

How concerned are you about the regulatory scrutiny the product could receive?

Galli: We haven't had any external discussions with regulators, but we've stayed close to our internal legal and compliance areas. The photo is anonymous, there's no personally identifiable information and we destroy it. Once you submit formal information that says you're interested, we keep all of those answers online. For that reason, we don't see any significant regulatory issues.