The impact of AI on Insurance Risk

Direct Line Group has always been a disruptor but with technological developments now moving faster than ever, we  stay at the forefront of innovation and continue to lead the market. A key part of this is understanding the impact of AI and Machine Learning on risk.

Technology is changing almost every aspect of our lives, and insurance is no exception; advances in tech are disrupting and revolutionising our industry across the board. And perhaps no-one is better placed to understand the importance of this than us at Direct Line Group.

In the mid-1980s, we were the original disruptor in the insurance market. We took the market by storm with a simple proposition: one product (motor) and one channel (phone). We had fully embraced InsurTech before InsurTech became a fashionable word.

Direct Line Group has always been a digital business, and today we provide insurance to over 10 million customers, across a broad range of products and through a wide range of channels. But we haven’t forgotten our disruptor status. With technological developments now moving faster than ever, the rate of change is asking us to look at what we need to do to stay at the forefront of innovation and continue to lead the market as it responds to the opportunities that are arising from the current technological revolution. And a key part of this is understanding the impact of AI and Machine Learning on risk.

Technology innovation is driving change in every area of insurance: from when we talk to you (customer engagement); how we talk to you (customer channels); and what we know about you (customer understanding); through to how we fix things (claims resolution); what we are fixing (autonomous vehicle technology/connected homes) and, of course, how we price risk. In all of these areas, we need to radically change the way we do business.

Let’s take one example: what happens when you have an accident.

Car manufacturers are introducing apps that are designed to help you through this process. They will, if necessary, alert emergency services, as well as identify immediate safety actions; they collect relevant information, such as photographic and other evidence – and may also store information about the traffic, weather, light and road conditions when the accident took place. This app will then notify your insurance company about the accident.
If we get to a point where we can then process that claim automatically, you may not need to talk to your insurance company at all.

Ultimately, we want to move from a ‘detect and repair’ service to a ‘predict and prevent’ one, and a good example of how technological advancements are supporting this is our capacity to insure a journey, as opposed to insuring a person.

Imagine you have decided to drive to Manchester for a few days. When we consider insuring you for this, we look at a number of things to assess risk: the length of time you have been driving; the make, model and age of your car; your overall car usage; and any claims and convictions. These factors are assessed as standard on an annual basis. However, we also now consider other factors, and the risks – and therefore the costs – associated with insuring that journey will vary depending on: the software level of the vehicle; your familiarity with the car (especially if you are using a car sharing service); the route taken and your familiarity with it; traffic conditions; time of day and possibly the time of year; who is in the car, and even what music is playing. All of this information is a ‘point in time equation’ – it is specific to this particular journey at this particular time.

We are moving towards a situation where, at the point of departure, your Integrated Vehicle Infotainment system can tell you the best time to go, what the route should be, even what music to play – and will provide you with the projected insurance costs of both options.

By starting to look at risks differently, we can get to a level of detail which enables us to give our customers more information about the risks that they are taking and how they can themselves adopt a more proactive attitude to risk. After all, it is not just about how easy we make the claims process, it is about how we stop you needing to use the process in the first place.

In part two of our explorations of the impact of AI and machine learning on risk, we will look at the impact of autonomous vehicle technology.