If AI can predict death, what about life insurance? – Opinion

If AI can predict death, what about life insurance? – Opinion
If AI can predict death, what about life insurance? – Opinion

A group of researchers from a Danish university developed a model called the “death calculator”, an algorithm to predict the stages of life until its end and which seeks to show the risks of commercial use of this data. According to the scientists involved in the study, the possibilities are endless, such as predicting health outcomes, fertility or obesity, or perhaps who would or wouldn’t get cancer. The model is based on anonymous data from millions of Danes, collected by the Nordic country’s National Institute of Statistics.

Then a question arose in the insurance market that does not want to be ignored: if death is predictable, what about life insurance? Will insurers be able to select customers and terminate contracts before the target date? Does natural death become a product of marked cards? These were some provocations that a reader made and we sought out some relevant opinions in the sector on the topic. After all, life insurance represents US$2.8 trillion in sales worldwide out of a total of US$6.7 trillion, according to a study by Swiss Re.

Nuno David, director of MAG Seguros, one of the longest-running life insurance companies in Brazil, is immersed in the topic of innovation, especially with regard to personal insurance. According to him, the availability of data that researchers refer to is much easier in a country like Denmark, which collects data in a continuous and consistent manner. “However, we have other research fronts that can change data from the past and, therefore, we have to follow the trends that feed and change this incredible genetic engineering”, says David.

One of the theses that the executive follows and tends to believe is that life insurance, with a focus on natural death, should be incorporated into health insurance, according to experts with whom he has routine contact. “Life insurance is heading towards a future, which we don’t know when it will happen, in which mutualism could disappear”, says the director of MAG, partner of the Dutch group Aegon.

Mutualism is one of the basic principles of insurance. It represents the contribution of several people, exposed to the same types of risk (mass of policyholders), to the formation of a common fund, made up of the sum of the premiums paid to the insurer. Now, what artificial intelligence (AI) allows is for insurers to create tailored products for their customers.

Roughly speaking, with the amount of data available, insurers will know, in addition to where they live, marital status, if they have children, what car they own, if they have fines and their name on the SPC, what people eat, if they practice sports, what medicines taken, whether they undergo preventive exams, stress level, travel and many other data that can determine the lifestyle and attention given to comprehensive health (physical, mental and financial).


Is this all bad? Not for people who take care of themselves. Given the power of AI, Nuno David believes that the creation, underwriting and pricing of personal insurance will be much more holistic, with insurers looking at customer health data from the past, but with the radar on current behavior. “It has already been proven in studies that a person can change their DNA with good or bad attitudes adopted in their daily lives. DNA tests are still for few, but they have helped to understand what future behavior will be like. This will be a condition for calculating the price of accidental life insurance, which should become the main product in the personal sector in the future, which still seems distant to me”, says the MAG executive.

Bernardo Correa Ribeiro, co-founder of Azos, insurtech which offers solutions for life insurance, agrees with Nuno David. He states that this type of analysis of the algorithm created based on Danish data is a “back test”. They took several old scans of people, created an AI model and tried to guess when the person died. Technologies, habits, diseases and treatment change a lot over the years. Therefore, a “back test” that got cases right decades ago would not have the same accuracy for future events.

“I am very skeptical that this could extinguish the product, but rather personalize life insurance and make it, in some cases, cheaper and, therefore, increase competitiveness favoring the end customer. I agree that there will be some cases of refusal by the insurer for having access to data and models that will indicate a more precise risk, but this does not prevent them from adjusting the price or offering coverage that would suit that client. So, I believe that this will encourage the market rather than extinguish it”, says Ribeiro.

Rogério Araújo, partner at TLG insurance broker, specializing in personal insurance, has a more traditional position. He argues that AI is not a threat to life insurance. “I believe more in AI helping us to awaken the needs of society, as a tool for financial and insurance education, than a threat to our business”, he says. According to the broker, even without the interference of AI, it is now clear that the purchase of life insurance does not depend on the money, but rather on the health condition of the insured person. “There is no point in having money and not being healthy. Without health, the consumer cannot buy a policy. Hence the need for insurance as soon as possible, although it is never too late”, says Araújo.

Veronica Martire, an educational consultant who has lived in London for more than 20 years, is an example of the change, which received a push from her life insurance company in England, where the relationship with the consumer is more mature than in Brazil. Although Brazilian insurers offer many benefits, few use them. “Just by stating in the health questionnaire that I do physical activities, I received discounts on life insurance. The insurance company offered me facilities to purchase a watch that tracks and encourages exercise and I accepted the challenge”, says Martire.

The more you exercise, the more points you earn. “Walking helped me lose weight and well-being led me to swimming. If I have a check-up, I earn points. And I can exchange points in a network of suppliers linked to quality of life. I have a discount, for example, when purchasing organic products. Result: I have already lost 12 kg in six months, consistently changing my lifestyle. I can no longer imagine doing something that puts my well-being at risk and, with each renewal, my insurance costs less, even though I am a year more mature”, she says.


When asked whether the sector is moving towards a selection of risks, offering insurance only to healthy people, Araújo is emphatic. “I don’t see the risk of a selection that restricts the acceptance of policyholders, since the concept of insurance is mutualism, and our regulatory entities would not accept this principle being disregarded”, he says. “And even if the AI’s assertiveness in predicting people’s deaths is proven, we will still count on the percentage of uncertainty. And I would pay to protect myself from the doubt of whether or not I was in the 22% error rate.”

But on three things everyone agrees.

The first is that AI is a great benefit that can help insurers in developing products, given the countless and diverse needs of society, as well as in pricing risks in a personalized way. “More assertive pricing, for example, at the time of a life insurance proposal interview, in which the insured’s life profile and health condition are assessed, allows for a more appropriate price, with discounts or increases in the fees charged. ”, says Araújo.

The second is that personal insurance is moving towards offering lifetime benefits to customers, such as compensation for the diagnosis of a serious illness, such as cancer, as well as incentives for practicing sports, such as discounts at gyms, natural food chains and even discounts for the purchase of watches that encourage practices that generate well-being.

And, finally, the interviewees understand that the benefit of buying an insurance policy, as a certain risk, as a way of composing or leveraging an asset, a reserve for succession planning or inventory expenses, is an essential tool in any financial planning, whether for the low-income population, breaking the cycle of financial deprivation for new generations, or for the high-income society, as a solution for financial intelligence, asset leverage and business succession.


Everyone claims that life insurance, even when predicted, is the only one with certain risks. In other words, the essence of protecting those we love persists, even if we know, approximately, when we are going to die.

The article is in Portuguese

Tags: predict death life insurance Opinion



PREV 3 healthy and simple recipes to make
NEXT US confirms second case of bird flu in humans