Artificial intelligence (AI) could revolutionise the insurance industry. 2018 has started with a multitude of similar eye-catching headlines, so I thought it was time to wade in with our thoughts. Is this statement fact or fiction?

Why the hype?

Let’s start with a definition. According to Wikipedia, AI is “intelligence displayed by machines, in contrast to the natural intelligence displayed by humans”. However, in the context of this article, Tom Davenport, co-founder of the International Institute for Analytics (IIA) and fellow at the MIT Initiative on the Digital Economy is more helpful. In his view “About 90% of AI has a statistical underpinning.” This statement underlines that, in essence, the concept is nothing new. So why the hype?

  • Technology:

    What is new and therefore generating all the excitement is the rapid advancement of technology. Nick Daffan, CIO at the data analysis firm Verisk Analytics says “We’ve been going backwards and forwards on AI for 15 to 20 years, but I believe we’re at a tipping point because of the advancement of algorithms, storage in the cloud and computing power.” It is here that we find the potential for insurers.

  • Big Data:

    The insurance industry creates an astonishing amount of data. Driverless cars alone create four terabytes a day of data according to Henry Burton, founder of Artelligen, a start-up aimed at bringing AI into commercial insurance. Computers are simply quicker and more effective than humans at analysing these levels of new data. “Where we see machine learning changing the industry is in how the world models and understands risk,” says Burton. “Without question, machine learning will replace statistical models. … Statistical models cannot handle that amount of data,”

  • Improved risk assessment:

    AI can ‘learn’ the business of covering risk. Logically, if you replace an underwriter with an algorithm – the algorithm is far less likely to make an emotional response than a human being, thus reducing the risk. We see the role of the underwriter evolving to a role where numerical skills are critical to apply the insights provided by the ever growing amounts of data into business logic. Underwriters will be needed to refine algorithms and will then be able to use the power of AI to improve the speed and effectiveness of risk assessment.

So the potential is there, but in our view AI will work alongside existing processes to improve rather than revolutionise. Bill Franks, the chief analytics officer at IIA believes companies will need to think about hybrid analytics if they want to remain competitive. Franks said “The architecture can be an insurance policy,” By this he means the technology used in the future needs to be flexible, making it easy to add or get rid of tools. In 2018, analytics professionals should start to consider how well their core insurance platform will integrate with the range of tools available.

2018, time to take action

Tech-savvy insurance start-ups are already circling. They see an old-fashioned industry, complacent and still sitting on fat profits made in the past. Lapetus, a US-based start-up is an interesting example. Life insurance traditionally involves an in-depth assessment of an applicant using a complex actuarial model. Lapetus, believes a “selfie” can replace much of that. Computers are programmed to scan a customer self-portrait and analyse thousands of different regions of the face. They are looking not just for basic information such as gender but for clues about how quickly the person is ageing, their body mass index and whether they smoke. This, together with other information collected from the customer, allows the computers to calculate what Lapetus says is a far more accurate prediction of life expectancy than traditional methods can provide. And the whole process takes only a few minutes.

This is evolution. The insurance process is basically unchanged. There is an application phase, including a new element of the self portrait. There is an underwriting process, albeit it electronically using an in-depth picture analysis and then there is the final risk decision. What Lapetus does represent is the potential to speed up insurance model using the power of AI to meet the digital expectations of the new market. It should also flag up a danger signal to the traditional Insurers. Technology introduces change that is inevitable and should not be ignored.

Revolutionary? Not at this stage. AI of itself will not revolutionise insurance but it has the potential to radically improve the industry. To revisit the views of Bill Franks, a good way to start is a review of the core technology. Without the infrastructure to benefit from the capabilities of AI, Insurers will undoubtedly lose out.