The insurance industry is looking at the means to develop new business models that rely on the mining of large data sets in order to identify customers, price risk and analyse claims. Not only does the application of artificial intelligence (AI) have the potential to reduce costs by reducing headcount, it also has the potential to improve the accuracy and speed of decision-making and transform business processes. Ultimately, the benefits of AI should be significant for the insurance industry’s customers.
What do we mean by AI? In broad terms, AI is the field of computer science that includes machine learning, natural language processing, speech processing, robotics and similar automated decision-making. AI enables machines to carry out tasks that would otherwise be dependent upon grey matter. There is a spectrum of sophistication of processes and this includes making decisions of various degrees of complexity.
Applications of AI might include the use of ‘chatbots’ to assist customers with insurance applications online and guide them towards tailored products and services. Over time, chatbots can learn from each interaction to enable them to provide better and more sophisticated products as they develop in a real-world environment. Machines can hold far more information about the product and its suitability for the customer than a sales agent can. Customer on-boarding, claims documentation and customer records can be easily stored, searched and analysed with minimal human interaction. This means existing insurance approaches to customer information can be transformed to deliver a much faster and more streamlined experience, often bringing new insights into the customer’s risk profile. Through Big Data analytics, data about the customer can be sourced from a far greater number of sources and analysed with limited or no human intervention.
Connected devices or the ‘Internet of Things’ enable insurers to get accurate risk information that can help the underwriter understand the risks that their customers face with far greater depth and accuracy as they can monitor the use of the insured asset (whether a life, vehicle or property) over the period of cover. Such devices allow insurers to develop a ‘scorecard’ of customers’ risk profiles, thereby providing more accurate pricing at inception and renewal. Claims handling can be undertaken without human intervention as evidence can be scanned, assessed and paid out (or denied) by machines. Trends in claims can be identified by AI in conjunction with Big Data analytics, leading to far more accurate risk management information and more effective fraud prevention.
All these new applications will rely on access to data. This might be customer data of varying degrees of sensitivity and might also be data gleaned from third party sources, such as social media or public records. In themselves, such data access and analysis would seem to be a good thing, as in theory they lead to better products for customers which can be purchased at a keener price.
Why, therefore, has Tesla and Space X founder Elon Musk called work on AI applications “summoning the demon”? Musk believes that, without regulation, some AI applications will be harmful to humanity.
So, what are the risks and how might they be managed by firms in advance of prescriptive regulatory requirements? Read our full article on AI and Insurance here.
 Elon Musk was interviewed in October 2014 at the MIT AeroAstro Centennial Symposium.