AI is still evolving, anyone who has used a SatNav on UK roads knows that. When the instruction “turn left” actually means a sharp bend, you can see that AI’s decision making process entirely depends upon the data being fed into the algorithm. Same principles underpin insurance quote pricing, admin and claims – you need accurate datasets before AI can truly replace, or even assist, humans.
So as ChatGPT is hyped up with the same fanfare as 5G, Cloud computing, Metaverse, or – if you’re old enough – Windows 98, it’s timely to ask where the big AI gains are across the insurance space.
Daniel Derham, Insurance Specialist, SAS UK & Ireland sees huge potential in understanding risk, at a much more granular level;
“AI & machine learning are a win, win for everybody in insurance. The use cases are really widespread. Starting with customers. Insurers should be able to offer the right product and policy to their customers so they are covered for what they need at a competitive price. Previously, the market’s approach was ‘here’s our product, this is what it contains and this is what you get – take it or leave it’. But now customers expect their individual needs to be met – an expectation largely accelerated by how the pandemic changed personal circumstances. If someone was suddenly at home five days a week and not in the office, they might think ‘why should I pay for car insurance covering 10,000 miles a year?’ They’d want their insurer to be accommodating.
“By using cloud analytics and AI to analyse big data sets, it helps insurers understand their customers and their risk profiles much better. It gives them confidence that they are charging the right premium associated with underwriting a specific product and the relevant risks for the needs of that individual customer. The right risk for the right price.
“This is what ultimately can make an insurer’s business more sustainable. They can offer tailored policies to improve customer satisfaction, retention and longevity – discouraging them from shopping around.
Matthew Biboud Lubeck, Amperity also sees economies of scale, utilising the Cloud, plus benefits from things like verifying customer ID using AI;
“Artificial intelligence has revolutionised the way companies identify, understand, and connect with their customers. While early iterations of AI were slow to meet enterprise-scale ambitions, new advances in AI and machine learning (ML) have unlocked capabilities once thought impossible.
New tools, made available in the last couple of years, are taking advantage of cheaper cloud computing costs and stunning integration capabilities. Businesses can now manage massive datasets and ingest raw customer data across every touchpoint — from online and on-site interactions to loyalty programs, email interactions, and finance systems — then use ML to resolve identities even when records lack unique identifiers across systems.
Banks and insurance companies are leading the way in terms of using AI to achieve a 360-degree view of their customers across all of their offerings. With a unified perspective, they are able to drive upsell and cross-sell more effectively within their owned channels and are able to reduce advertising by removing the possibility of trying to reacquire the same customer; it is important to remember that the successful use of AI relies on solid data foundations that empower companies to use data to deliver more personalised customer experiences.
Enriched and unified customer profiles, made possible via the strategic use of AI, connect essential customer information, including demographic, loyalty, browse history, email engagement, and product purchase data. This allows companies to generate richer and cleaner features that, in turn, improve the ML model performance.”
It’s worth noting the benefits of AI when looking for ptential fraud. Companies like Percayso-Inform and LexisNexis for example can overlay custom fields of data, so that you can see things like linked addresses, previous claims per individual, credit applications or debt, particular locations which are theft attempt hotspots and so on. To do such tasks manually would take days, but AI enables the sifting process within seconds.
Daniel Derham from SAS UK& Ireland has these comments;
“Taking customer datasets of activity, when analysed on a huge scale through AI, is also enabling insurers to combat fraud – by ultimately helping them to spot patterns of fraudulent behaviour and prevent fraudulent claims being processed. There are many insurers who are using or considering AI models to do this, given that fraudulent claims continue to rise.
RESPONDING TO CONSUMER DEMAND
“Looking internally,” adds Dan, “AI helps insurers to forecast resources much more efficiently. If for example an insurer has a full call centre Tuesday to Thursday but data shows that actually most calls are coming in Monday and Friday, they can plan better and optimise shift patterns so lots of agents aren’t sat at desks when call volumes are low.
“As in many sectors, the customer used to be two or three steps removed from what’s good for an insurer’s business. Now everything is led by customer demand and because of that competitiveness, insurers are having to use every tool at their disposal to offer a better product and service. Those who aren’t leveraging AI are already getting left behind.”