AI, blockchain and the future of insurance

by Nadeem Shaikh

This photo, and hundreds of others like it, are becoming defining images of our age. Mexico beach in Florida was destroyed by the impact of storm Michael. This house remained standing while its neighbours were washed away. As the insurers move to assess the damage of this season of storms they are increasingly turning to artificial intelligence to make their assessments. In the future of the Fintech industry it is the use of AI and blockchain that will define those products that breakout from start ups to become market leaders. The process of thinking about the application of these new technologies begins with thinking about this picture.

There are three ways in which AI can help the insurers and the insured in dealing with the aftermath of this terrible event and can make judgments based on it to inform financial decisions in the future. First, AI can look for patterns.  Second, AI can interact with customers.  Third, AI can use Blockchain to change the way insurance works.

AI can explore the vast amounts of data that are collected on people every day. If you think about your retail transactions, every single thing you buy in the supermarket becomes a data point about who you are as a person. Retail has fully embraced this technology. Advertising is tracking your online shopping to reinforce and seek another purchase. Your personalised discount vouchers echo your purchase history. This is now being increasingly applied by the insurance sector. The search for patterns is obviously useful in fraud prevention. If transactions on an account suddenly change their pattern, machine learning can spot that pattern change and respond. Patterns in weather are obviously also going to make a significant difference to insurance dividends in the future. The New York Times reported: “After Hurricane Andrew, a Category 5 beast, ravaged Miami-Dade County in 1992, new construction in the southern portion of the state was required to withstand 175-mile-an-hour winds. In the coastal Panhandle counties affected by Michael, the requirement is lower, for 120 to 150 miles an hour, and the rules for certain kinds of reinforcement have applied to houses built more than a mile from shore only since 2007.” The people who built this house, built it to withstand winds of 250 miles an hour. Weather patterns are changing so rapidly that it may only be AI that can keep pace and keep insurers and the insured informed. These patterns will change much more quickly than building regulations will adapt.

In addition to these kinds of assessments, AI can look for deeper patterns in risk management and increasingly in predictive analytics. This has long been used in algorithm based trading, and there will always be room for more Fintech innovation in that space, but now the use of satellite imagery linked to big data sets is making an entirely new generation of commodity trading tools more widely available. In turn, looking for patterns in these daily and sometimes hourly images that have a zoom capacity down to 41cm, can spot building regulation violations, absences of planning permissions for new builds and movements in water bodies that indicate changes in flood threats. Drones and the Internet of things also provide rich fields for exploring the collection of big data sources. Fintech innovation in insurance and other areas is using all of these tools and more.

The other dimension of AI is the interaction with customers. So many simple financial transactions are routine, repetitive and functional, chat bots, automated voice responsive systems, can be used to handle more and more of them. As natural language processing advances, these systems should become better and fintech innovators are working hard to make them more responsive and human like in their interaction. These advances are absolutely essential in the use of AI in marketing from direct calling to the algorithms that place adverts on the goggle search pages we use.

All of these features come together in the application of block chain technology to the insurance sector through three mechanisms that combine elements of all of the above: Smart Contracts, Smart Assets and Smart Execution. For most people, most of the time, insurance was something running in the background and the contracts on which it was based were passive tools. They might come into play at the end of a life or in the rare case of a flood but otherwise they were usually hardily looked at and frequently never read by the customer. Today and in the future, we need Smart contracts that can adapt to changes in circumstances and use consensus-based logic to agree on the meaning of those changes. As climate changes, for example, a Smart contract can reflect changes in circumstances. As banking and insurance opens up, and consumers are able to compare prices between providers much more easily, Smart Contracts could develop the capacity to be price comparison vehicles. In turn, Smart Assets might develop into virtual representations of at risk property that spread their insurance across a range of providers to maximise cover and the possibility of claims being awarded. Finally, when claims come in, especially in mass incidents, that will only become more frequent as climate change worsens, Smart Execution, the fast efficient transfer of data to make assessments and then rapidly make the right payments or arrange the right services, will become the norm.

The possibilities are endless. The Fintech industry might have been slower than retail to realise the opportunities of AI, but in areas like insurance, they may have no choice but to embrace innovation with both hands or be left behind because the application of AI to blockchain allows for completely new entrants to these markets who understand how to use these technologies to stay one step ahead.