Artificial intelligence (AI) is the largest single disruptive technology impacting InsurTech. I work with AI every day and know that you can run, but you cannot hide from the fact that AI is going to change the insurance industry fundamentally.

While it’s intuitive that AI can break the paradigm of how users interact with our systems, I want you to imagine with me what else AI technologies can do for InsurTech. Selling insurance and settling claims are the easy parts. AI is going to do far more, it’s going to give us new kinds of insurance, new ways to look at risk and allow us to look at data at a scale far beyond our human limitations.

Cognitive Computing

Cognitive computing – computers that understand the intent of conversation – will own the future. These computers will also be imbued with the ability to understand that we’re human and have sentiments and personalities. Cognitive computing will give us unprecedented advances in automation that will provide non-stop services from sales to claims. AI’s most powerful feature will be the system’s ability to understand who we are as customers, what we do with our lives, and what insurance we need. The future of insurance isn’t depersonalized as some may fear, in fact, AIs are going to make it more personal than it has ever been before.

Chatbots Change how Humans Interact with Machines

Chatbots are the hottest topic in InsurTech with good reason; they are the first real technology that changes the paradigm of how humans interact with machines. Chatbots are designed to be the interface that gives technology a voice so that we can talk or text in natural language with a machine. It is a foundation technology upon which more sophisticated automated advisors will eventually be built, so ignore chatbots at your own peril. You probably couldn’t ignore them if you tried, as they are already in our living rooms and allow us to play music, turn on the lights, or buy detergent when key words are detected. InsurTech chatbots are coming online now that will allow us to buy coverage, check coverage features, and do many of the simple jobs done either by FAQ sections of the Internet or call centres.

While chatbots are still a long way away from having a full-on conversation with a human, the simple tasks they perform can bring dramatic increases in efficiency and significantly alter how we interface with insurance platforms. They provide an instantly scalable interface that can be seen as an intermediary between a call centre and a website. Most importantly, they allow 24/7 access to information in a delivery mechanism that can be rolled out and integrated into insurance platforms in months. They will be the constant companion of your customer and allow you to serve them better throughout their customer journey. The most significant benefit is that they can be configured to help educate the user and allow them to do more by themselves. Think of this as AI giving humans greater capability to handle their own affairs rather than having to wait for a human agent to act as an intermediary. Does this scare you? How many of you still want to call your airline and talk to an agent to select a seat rather than booking it on your phone with an app?

Advances in chatbots now allow the “bot” not only to converse with us, but to determine the tone of the discussion through a process called sentiment analysis. This is where bots become personal in their responses as they can detect the client’s emotional state and deal with them according to their level of satisfaction. If chatbots let you communicate with a computer, sentiment analysis is what will give the computer a heart. In the most simple use case the bot can determine when the client isn’t satisfied with the service provided by the bot by understanding the tone of their response. Just as the user starts to sound frustrated the bot can volunteer to transfer the call to a human agent. In more complex uses, sentiment analysis can be deployed to see how users are interacting with a new system. You can actually compare the users’ sentiments to the prior version and get unambiguous results of customer satisfaction with system improvements.

Even if we integrated chatbots with a heart into every part of an insurance company it still would not resolve the problem that we continue to sell product by silos such as health, auto, and life. This is hardly the view of the future that any of us have but we seem stuck to this view of the world because, for now, our products are each covered by a unique contract. What this fails to do is to relate to the customers based on their needs or goals.

Artificial Intelligence and Goal-based Insurance

Goal-based insurance is a way of trying to understand customers’ needs better based on what they want to do or achieve in the year ahead, or potentially with their lives. If you ask a customer whether they need supplementary foreign medical coverage you are likely to get the answer “no”. If you ask the same customer if they plan any wonderful holidays abroad this year you will probably get a “yes”. Customers tend to think in terms of what they want to achieve, or do, rather than which silo of insurance is applicable to them. Goal-based insurance is a means of mapping out the customer’s life journey in a way that uncovers potential areas for insurance rather than probing the need for individual policies.

AI and goal-based systems are the perfect combination of technologies for creating a major disruption in how we sell insurance. Goal-based systems are just now being developed that “gamify” the process of understanding a client’s goals. Gamification is a way to simplify the process of understanding the client’s goals without frustrating them by asking tedious questions. First-generation goal-based systems are going to use chatbots alongside gamification to make the experience interactive. Second-generation systems will unleash the full potential of AI, vastly simplifying the process so that the system only needs to ask two questions: “Who are you?” and “What can I do for you?” AI connected to big data will do the rest. It will understand who you are as a customer, where you are in life, and your optimal insurance coverage based entirely on analysis of your big data footprint. It will then have a discussion with you to refine the coverage better and provide optimal insurance just for you.

Sound far-fetched? First-generation chatbots are now in service at a number of insurers providing customer service that never sleeps and the first chatbot selling travel insurance on Facebook Messenger was launched in Singapore in 2017. While most of the bots built right now will not fool you into thinking that you’re talking with a human, don’t dismiss the simple things they may do like answering questions or helping guide customers through sign-up or claims processes. It may seem like the leap between first- and second-generation systems that know about you is a big one. It isn’t, the technology is actually already here, the problem is gaining access to data sources that link you unambiguously to your data. This is changing quickly since governments spurred by “Know Your Customer” (KYC) regulations are putting directories of citizens and taxpayers online to reduce fraud. Singapore is launching its KYC system for Singapore residents in early 2018. Once these KYC systems come online, the personal data can be crossed with social media platforms to provide you with everything you need to put together a basic customized insurance program. We’ll see the first of these systems by 2022/3.

If AI is going to help make policies more personalized, the natural fallout is that insurers will need to learn to blend our policies into a more goal-based view of coverage rather than selling a series of disjointed policies. Successfully underwriting these new policies will also require AI as they will be priced upon multiple dimensions of risk characteristics that will be taken from the client’s actual behaviour. Take auto, for example, where each client will have a customized risk spectrum based on five to ten dimensions, each with a differing risk score. It’s easy to imagine a risk spectrum composed of days used, miles travelled, region, speed, telephone usage, passenger numbers, braking force, and any other data on-board telematics may give us. Now take this auto profile and add it to similar metrics collected on the insured’s health, property, occupation, and hobbies and you can build a policy that gives your clients more personalized coverage than ever before. Just as importantly, you’re going to reallocate your clients’ money to areas where they need the most protection and better underwrite their “real” demonstrable risks. It’s a future that I look forward to.

The Next Generation of Insurance Will be Built on AI

So far, we’ve only considered what AI is going to do to our existing systems to improve the customer experience and access behavioural data more easily. The next step for AI will be to help us design new types of insurance coverage for risks that can only be defined with use of AI. The most obvious place for AI to do this is in cyberspace since humans aren’t equipped to monitor or respond to threats in cyberspace until after the damage is done. Cyber security is really a new frontier for the insurance world and an area where AI can monitor, mitigate, and quantify risks faced by our customers. They can scan our cyber-life for signs of malicious attack, raise alarms, and give a clear unambiguous view of the severity of the attack or even potential damages. The threats monitored by AI will be subtler than data theft or loss due to malicious activities currently covered in cyber policies, but go to the root of insuring our cyber-life and that this life has value beyond the data. Access to granular data on health, life, loss, driving behaviour, and personal habits will shatter actuarial norms. Just as significantly, AI’s access to new data from our cyber-life will give us new quantifiable dimensions of risk that we can price and manage.

I know some of you are thinking that this is going to happen in the distant future and that it won’t affect you. You’re mistaken. These changes are happening now. AI is going to make insurance easier, cheaper, more personalized and will manage more kinds of risks than we can currently conceive. It will change how we interact with our customers and improve the overall customer experience. Paper and PDF forms are the cement and bricks that built our industry. The next generation of insurance will be built on AI. So start thinking more like an AI company and less like an insurer if you want to survive this paradigm shift.