AI in insurance forces big changes to traditional industry

With new technology come new risks

The fundamental reason that insurance exists to begin with is to provide individuals and companies with a way to cover the financial losses that arise from unexpected events. Insurance only makes sense as a business, however, by making sure that the likelihood of those events and the significance of the losses are balanced appropriately. So insurance firms not only need to know ahead of time what sorts of risks businesses will want to cover, but also how common those risks are and how easy they are to prevent.

Looming large in the windshield of insurance companies is the increasing use of autonomous vehicles. Insurance companies are already struggling with how to deal with liability in the age of ride-sharing services such as Uber and Lyft.

Now, insurers are faced with new challenges, such as providing insurance coverage when the human behind the wheel might not actually be in control of a car in self-driving mode, or when the company who owns the car -- such as Uber -- might not be the company that made the car -- such as Volvo -- and the company that made the car might not have made the autonomous technology. Insurers will have to consider these new challenges to assign liability, assess risk and determine loss ratios.

With the emergence of autonomous trucks, aircraft, ships, trains and drones, insurance companies are facing unknown risks, liabilities and costs. Either the autonomous vehicle operators will have to run their businesses without insurance or they will have to find some way to bring insurance companies up to speed on how to properly assess risk and loss.

Autonomous vehicles are not the only new threat vector posed by AI. AI is enabling new forms of cybersecurity threats with subsequent risks for individuals and businesses.

Cybersecurity and cybertheft coverage are two of the hottest growing areas of insurance coverage for businesses. Companies and individuals are increasingly interested in protecting the losses that might come from AI-enabled fraud and technology-enabled personal injury. In the world of fake videos, fake audio and fake images, there's a whole new range of liability for individuals and companies.

Another challenging issue currently is insuring liability that companies might have when using AI-enabled products. Will insurance companies cover business losses when an AI system inappropriately performs some action or causes some harm that a human would not have otherwise caused? Will simply using AI at all in the delivery of a product or service cause insurance companies to deny claims because of the unknown risks and liabilities introduced by machine learning and AI systems? The answers are unclear.

Using AI to better manage risk

Increasingly, insurance companies are eyeing the opportunities they can create by applying AI to their own operations and products. Many automobile insurance companies have seen significant benefits using telematics devices that plug into cars to assess the risk posed by their insured drivers.

For example, Progressive has collected over 14 billion miles of driving data with their Snapshot device. This device is helping the insurance company better understand the risk profiles of their drivers, rewarding the safest drivers and appropriately managing the riskier ones.

AI in insurance can provide significant value by extracting key patterns from the vast amounts of data collected by insurance companies as part of their policy, customer, claim and risk data. Healthcare insurers in particular are seeing huge benefits by spotting early signs of health issues that can be addressed with better and less expensive treatment than would be available later.

A variety of machine learning and AI companies are specifically applying their products to the insurance industry to help them better identify and manage risk and risk calculation, improve underwriting and risk prevention, manage fraud, and provide right-sizing of insurance coverage for their customers. In many ways, AI provides augmented intelligence to insurance companies, enabling them to be more efficient and effective when delivering their insurance offerings.

AI improving the customer relationship with insurance companies

In addition to improving insurance operations, AI in insurance promises to improve the relationship between insurers and their customers. Already, the use of image recognition and natural language processing is being used to speed up claims processing.

For example, many auto insurance companies are using smartphone camera images to provide instant loss assessment information and health insurers are using handwriting recognition and image recognition to increase the speed at which they handle health-related claims, invoices and receipts.

Because insured losses can happen at any time of the day or night, insurance companies must have claims processors available 24 hours a day to handle unexpected claims calls. Many insurance firms are already implementing text and voice chatbots across a wide range of platforms to handle claims and customer support in a reliable way, even when massive loss events such as natural disasters, terrorist-related incidents, mass casualty and damage events happen. No one wants to be on hold with an insurance company when you experience great loss and have your life to piece together. AI-enabled conversational assistants can make the insurance interaction smooth and seamless.

There's no doubt that AI is impacting every corner of our business and personal lives. Insurance companies are surprisingly at the forefront of those changes. The insurance companies that can appropriately manage the new risks and take advantage of the new opportunities will continue to thrive and operate as they have through every major technology transition that's happened to date.