By Eleanor Brodie, Sr Manager, Data Science at LexisNexis Risk Solutions

Artificial Intelligence (AI) and Machine Learning (ML) are being used in many walks of life today, but insurance has to be one of the most exciting. Almost everyone needs to buy insurance and some of us will have the misfortune to find we need to make a claim. At a base level, AI and ML are helping to make those experiences better.

Consider the volume of customer data held by insurance providers in claims, marketing, underwriting – often held in different silos, sometimes in different sub-brands. AI and ML techniques can help to make sense of that data. The linking and matching technology now available to the insurance market means a single, consolidated view of the customer can be created based on all previous touchpoints. This gives insurance providers a more holistic view of that individual to best support their needs and provides a strong foundation to build that picture further by bringing in additional data sources.

Insurance providers need to operationalise growing data volumes

These additional data sources include market-wide contributory data – policy history, past claims, quote history – some of this data has been gathered for over 6 years. Add in public address and ID validation data and other external data sources, for example environmental data, property characteristics and data related to assets such as the safety features on a motor vehicle or the data from a telematics device. Insurance providers need to operationalise all of this data, bringing it in at the right time to support quotes, price a risk, expedite a claim, flag possible fraud or understand a cross-sell opportunity.

ML helps take the guesswork out of home insurance applications

AI and ML techniques are making that possible, enabling data based decisions to be made, at speed.  A great example is the way the application process has been simplified in household insurance through ML techniques, improving pricing accuracy while cutting the time it takes for a customer to gain a quote. Prefill and data validation solutions speed the whole process but are only possible through a huge amount of modelling, linking and AI-ML techniques to pull all the data together to return accurate and up-to-date information on the person and property. For the customer it means no more guessing at rebuild costs or property age.

Eleanor Brodie

AI can deliver location intelligence for businesses

For businesses, AI can also provide valuable insights regarding a potential location for a new branch or business relocation – footfall, crime rate, exposure to perils or other local circumstances that increase risk in commercial property insurance. Armed with this valuable knowledge, the customer may take preventative measures which has the benefit of reducing the risk and potential claims costs.

Making IoT data meaningful

With the increasing volume of data coming from connected things, data normalisation through ML techniques is creating standardisation and consistency for usage based insurance based on this data.

Whatever the source of data – aftermarket telematics devices, smartphone apps, connected vehicles, even in the future from smart home data – data normalisation means consumers can enjoy an improved shopping experience using the consented data from their device or vehicle and insurers have consistent quality standards and consistent pricing for all consumers. Individuals then benefit from being judged based on their individual behaviours as is already the case in telematics insurance, rather than paying premiums based on average habits.

Helping insurance providers understand ADAS fitments

A prime example of data normalisation is behind a new solution[i] to allow insurance providers to price based on the ADAS features on the car, at a Vehicle Identification Number (VIN) level.  Machine learning has been used to scan millions of lines of car manufacturer vehicle data to logically sequence and classify vehicle safety features and component’s intended operation or purpose to create an ADAS classification system. This task would have been extremely difficult, time consuming and error prone without the use of AI/ML.

Taking the pain from claims

Motor insurance claims are also benefiting from AI/ML techniques as virtual claims handling speeds up claims resolution, cuts costs and provides a smoother customer experience.  Image recognition technology captures damage or invoices, runs a system audit, and the claims is paid automatically if it meets the right criteria. Bringing in historical policy and quote history to claims in the future may add an additional level of security prior to an insurer releasing any claim payments.

Many customers with telematics policies benefits from an improved claims experience thanks to AI and ML. From the point of impact through to claim resolution, telematics data can allow insurance providers to get on the front foot at first notification of loss (FNOL), to support the customer post-accident with emergency services, roadside recovery, vehicle rentals and repairs whilst providing invaluable insights regarding the circumstances of the collision.

Insurance providers can look at a range of data such as air bag deployment, impact sensor activation and g-force metrics to understand claim severity and bodily injury potential. They can also bring in vehicle build data to understand the repair cost and potential impact to expensive ADAS features.

Finding new data points to help predict risk, for use in pricing

In pricing, machine learning algorithms can expedite the identification of the most predictive attributes behind claims losses for use at point of quote. The most recent data points are cancellations data and gaps in cover in motor based on industry contributed policy history data.  Insurance providers can also distinguish between policy changes directly related to the first national lockdown versus changes that occurred outside of that time to support the fairest treatment of customers when they come to buy or renew their next motor annual policy.

Building insight from insurance specific behaviours and bringing in more data about the asset such as the car or home, where relevant can help insurance providers become more competitive, match their risks to the most appropriate pricing strategies and write the risks that meet their underwriting appetite. In turn, customers get more personalised quotes based on their unique risk characteristics across any line of business.

The insurance industry is being transformed through its use of internal and external data to better assess risk, price more accurately, reduce claims losses and support customers at every touchpoint.   At the heart of this transformation is analytics, Artificial Intelligence (AI) and Machine Learning (ML).

From making the insurance application quicker and simpler to expediting the claims process, AI and ML techniques are creating new insights, allowing human skills – the ‘personal touch’ – to come in at the points it matters most to the customer.

[i] LexisNexis Vehicle Build – https://risk.lexisnexis.co.uk/about-us/press-room/press-release/20200618-vehicle-build-uk