Event Recap: "Revolutionizing Risk: How sensors, Ambient Data and AI Are Reshaping the Insurance"

On September 5th, 2023, Prifina hosted a panel discusion "Revolutionizing Risk: How Sensors, Ambient Data and AI Are Reshaping the Insurance."

Among the panelists were:

• Prifina's CEO Markus Lampinen,
• Signe Amdi Holm, Co-Founder & CEO of Datasolvr, an insurance data technology company,
• Ashley Greenwald, Senior Associate from Huntsman Architectural Group and 
• The panel was moderated by Boudjemaa Nait Kaci, a financial services expert from New York Life Insurance Co. 

The panel discussion explored the current trends in the world where sensor-generated data and AI-powered applications permeate every aspect of our lives. Could such data be used to personalize insurance services and unlock new solutions to help us reconsider how we approach risk? 

The event on the implications of sensors and data on insurance gathered much attention from various representatives of different industries. In the following post, we review some of the discussion topics and provide a list of innovative solutions for insurance services.  

1. Paramount role of data for insurance 

The starting premise of this event was the realization of the huge amount of data generated by users themselves (e.g., data from sensor-equipped fitness trackers) and ambient data from sensors in office buildings and public spaces. Yet, appr. 80% of such data lies idle, unutilized.

Markus Lampinen from Prifina opened the panel discussion by explaining the paradigm shift that is happening in the personal data market - a new data framework the value from data is captured on the users' side (not locked away in centralized silos): 

“In the future, we want to see individuals getting the value from data captured across different data sources. We want to empower individuals with different types of applications that you own and that you run on top of your own combined data. Very simple things that help you sleep better or buy the right shoes, etc. But we already have those. The big picture? In 5 years, as an individual, I’d like to be able to predict what movie I’d like to see, and what better choices I should make.”

Ashley Greenwald from Huntsman AG explained the concept of "ambient data" - data from sensors embedded in spaces around us (temperature, noise, occupancy, vibrations, exposure light, humidity) - all of this data could be used to improve the efficiency of space management and human experiences. Such data from sensors in buildings is particularly interesting for building owners and renters.

2. Tight Regulations Limiting the Use of Personal Data

Signe from Datasolvr and Markus discussed the fact that data privacy regulations and well as risk-awareness of service providers limit the use of individual-level data for personalization. Companies don't want to face data discrimination lawsuits and opt not to use data in their services at all. While the individual-level data could be particularly interesting for creating different pricing models, data privacy regulations are tight both in Europe and in the US. 

3. Building Predictive Health Care System with Data

Participants of the panel discussed the ways how user-generated and ambient data could be used to enable predictive health and wellness services. This would be a major juxtaposition to the current "reactive" medicine which is extremely costly.

"In reality, when you go the hospital, you realize that many health-related data could have been avoided if various data sets were caught and utilized earlier." - Markus Lampinen

Another truth of the fact is that we are running short of doctors, physicians and experts: there are appr. 10 million doctors in the world for a population of 8 billion people.  And we will never have enough doctors. At the same time, people are increasingly eager to extend their well-being and the quality of life.  

That's where sensor-generated data and data-powered AI applications become increasingly appealing:

"Harnessing data from devices like the Apple Watch could help increase individual awareness of stress and potentially prevent harmful effects using predictive AI. That's one of the biggest opportunity for insurance service providers who ultimately want to mitigate risks." - Sille Amid Holm (Datasolvr)

​Individual-level use cases. The panelists explored several examples of utilizing user-generated data from wearables in insurance. For instance, renewing term life insurance is complex, with current pricing influenced largely by an individual's health condition, such as whether they smoke. For instance, smokers typically pay at least twice the premium compared to non-smokers of the same health. However, the potential use of comprehensive data collected over years could reshape this model. If data shows an individual quit smoking and improved their health over a decade, they might deserve a reduced premium, even if they're older. This concept suggests a more dynamic and responsive insurance model, though the decision rests with insurance companies.  

Group-level use-cases. Group-level data from wearables as well as ambient data could be a powerful tool to innovate in the insurance services market. Imagine, if an employer could prove certain habits of the employees (e.g., employees chose to walk up the stairs and sleep on average 6.5 hours a night) to the insurance company, this could offer additional insights and price adjustments for various insurance plans.

4. Creating New Consumer Experiences with Data

The panelists discussed various opportunities and challenges when it comes to building new consumer experiences with data. Many consumers need super simple interfaces and huge incentives; there is also a huge learning curve ("data literacy"). A huge part of this relates to the experts working in the User experience/user interface domain: How do you show the value from data to the individual? How to help people to better understand what to look for in their data?

There're many interesting insights about how people have adopted wearable devices: e.g., we see how much more aware about their biometric data people have become. 

However, it is quite difficult to build data-driven experiences in the insurance services market. The customer onboarding and customer journey has to be very different in insurance. Since more people interested in insurance services tend to be senior, the experience must be simple and understand. Here are some examples of gimmicks that could be used: 

• Savings notices: there can be an icon on a website which says that you can save xx%  with certain data
• Major rewards: Sille gave one of the most notable examples from Denmark - an advertising that looked like that "If you change your insurance provider, you will get a free car!" 
• Gamification & Incentives: insurance companies could use gamification strategies to nudge people adopt certain lifestyles; data could be used to offer better insights and rewards (premium discounts, points, gift cards, etc.) 

Accelerating innovation. The flip side argument of that is whether big players in the insurance market would like to innovate, or whether they'd rather prefer the status quo. Markus Lampinen from Prifina suggested that such a hold-out strategy by incumbents may not last too long especially when more innovative newcomers in the insurance market come up with more personalized and compelling products. 

5. Personal AI agents powering a new line of insurance services 

The panelists shared their common understanding that in the near future each of us will have our personal AI assistant experts (doctors, coaches, nutritionists, shopping agents). This world becomes possible as individuals get more ownership and control over their own data.

More generally, panelists agreed that user-generated and ambient opens new opportunities for virtually every stakeholder in the market, including innovation opportunities for insurance services that will become very personalized and specific.

Future vision. Imagine if your personal AI insurance agent negotiates on your behalf with AI-powered agents of insurance providers to find the best deal for your specific condition. The panelists even explored possible implications if an individual's own AI agent was aware of individual's DNA data and historical medical records  and how this degree of data accuracy and automatization could transform the whole understanding of risk. ​

We want to thank all the panelists and guests who attended the event and stayed after the panel to network and continue the conversation about the emerging personal data ecosystem.