The growing maturity of artificial intelligence (AI) and machine learning (ML) is changing the consumer’s experience of car insurance for the better by enabling agile new players to come to market with simpler, faster, more transparent and more affordable insurance products. This trend promises to change the face of the insurance industry over the next decade.
That’s the word from Alex Thomson, co-founder of AI-driven car insurance provider, Naked. He says that the maturity and power of AI and ML are already enabling car insurers to automate business processes, elevate efficiencies in the back-office, create innovative products and solutions, and transform their customer experience.
Thomson says that AI is a group of technologies that allow computer systems to sense, comprehend, act and learn in much the same way as a human might. The benefits of such technology are that it allows insurance providers and other organisations to automate many tasks and processes that may have required human intervention in the past.
AI encompasses a range of tools and technologies, including vision technologies such as facial recognition and analysis of expressions, speech recognition that can convert text to voice and voice to text, natural language processing that can understand contextual human speech or text, and reasoning algorithms that can create predictive models and anticipate behaviour.
Startups lead the race
Thomson says that startups are leading the global AI race in insurance because it is challenging for the incumbents to reinvent their business processes, reshape their organisational structure and clean up their data sources rapidly and thoroughly enough to put AI at the centre of the business. “To benefit from AI, you need to start by designing for AI,” says Thomson.
“AI and ML are still in the early phases of adoption, but they are already bringing enormous benefits to insurance customers,” says Thomson. “For example, we can offer our customers lower premiums because we use ML, AI and intelligent automation to keep our costs under control. We have an automated quoting system, so we do not need a large back-office infrastructure to manage sales and distribution.”
When a consumer seeks a quote, a friendly chatbot will help him or her through the process by asking a series of natural language questions. In the background, a set of algorithms will assess the consumer’s risk and come up with a quote in 90 seconds. “This is a revelation in an industry that has lagged other consumer-facing sectors in how it interacts with customers,” says Thomson.
“AI offers us the opportunity to remove much of the friction from buying insurance – a process that once involved filling in reams of paperwork, holding for a call centre or spending an hour in consultation with a broker. Now, it’s as simple as answering a few questions on an app or on a website.”
Reducing fraudulent claims
Some of the most powerful applications of AI and ML in insurance happen behind the scenes, yet have enormous benefits for consumers. Many insurers that are at the forefront of adopting technology now use predictive modelling and other AI tools to pick up discrepancies that may indicate that the policyholder has submitted a fraudulent claim.
This plays a key role in keeping insurance cover affordable, since as much as 30% of car insurance premiums in South Africa are estimated to pay for fraudulent claims. “In addition to highlight possible instances of fraud, the automated process allows for legitimate claims to be settled faster – often within a matter of seconds,” says Thomson.
“We believe there is still much more we can do to streamline the claims process using AI and ML,” says Thomson. “One example is using photographs and videos in automated damage assessments after a car accident. The more advanced solutions of this nature are faster, more accurate and produce more consistent results than a human assessor.”
More disruption from IoT plus AI
Looking ahead, Thomson says that the combination of AI with the Internet of Things and big data sources like traffic and mapping services could enable insurers to offer even slicker and more personalised services to their customers.
For example, a vehicle telematics system paired with a virtual assistant in your smartphone could offer real-time coaching to help you drive more safely. It could, in future, direct you away from high-risk routes when you input a destination into your GPS or allow for sophisticated per-kilometre pricing, based on where, when and how well you drive.
“We expect AI to change insurance in the years to come as much as Netflix has transformed TV or Amazon retail, and for many of the same reasons,” says Thomson. “AI enables insurers to offer immediate, transparent, personalised and user-friendly service – all terms that were not often used to describe insurance in the past.”