How is AI transforming the insurtech sector?

AI is fighting customer frustrations and driving efficiency in insurtech, explains Nick Martindale, but it should augment, not replace, humans

Artificial intelligence (AI) is impacting almost every industry, and insurance – and the insurtech sector on which it depends – is no exception, with applications benefiting both customers and insurance firms themselves.

From a customer service perspective, the use of chatbots is helping to answer queries in a more efficient manner, providing customers with instant answers around the clock, says Quentin Colmant, CEO of insurtech firm Qover. “AI-powered chatbots can assist customers with contract management, freeing up human agents for more complex issues,” he says. “Additionally, AI analyses vast amounts of customer data to personalise insurance recommendations. This allows insurtechs to tailor products to the specific needs of customers, ensuring they are presented with the most relevant options.”

Generative AI is tackling customer frustrations

The emergence of generative AI is likely to see this evolve further, using multiple data sources to provide even more personalised digital interaction. “General information typically provided through static and dynamic FAQs are likely to be superseded by a more interactive human-style chatbot, which was on the increase even before the advent of generative AI,” says Tony Farnfield, partner and UK practice lead at management consulting firm BearingPoint. “The ability to link an AI bot to back-end policy and claims systems will scale back the need for human intervention.”

Generative AI can also help target specific areas of frustration for customers, says Rory Yates, global strategic lead at EIS, referencing its own client esure Group. “They focused on a key customer frustration when calling a contact centre, which was repetition, so being passed from one person to the next, and needing to re-explain the reason for making contact,” he says. “Their use of generative AI helps alleviate this. Then at the end of every call, generative AI is used to summarise the notes, capturing the details of the call, making sure accurate records are kept.”

Internal efficiency is another major benefit of the effective use of AI. Steve Muylle, professor of digital strategy and business marketing at Vlerick Business School, gives the example of AI helping insurers to generate accurate quotes almost immediately. “In 2019, Direct Line launched Darwin – a motor insurance platform that uses AI to determine individual pricing through machine learning,” he says. “This approach has translated into better customer reviews and improved customer service.”

“Another example is in Asia, where insurance companies work with Uber,” he adds. “After an accident, insurers can ask nearby Uber drivers to check accidents, leveraging their knowledge of cars and their ability to take photos or videos for reporting, which can then be analysed by AI. This provides the insurers with more data, potentially from a third party, and is also a side gig for the Uber drivers.”

Another application is in the onboarding and training of employees. “AI-powered virtual assistants can guide new employees through the onboarding process, providing support and answering questions around the clock,” says Christian Brugger, partner at digital consultancy OMMAX. “Interactive AI-powered tools, such as virtual reality and augmented reality, can offer immersive training experiences, simulating real-life scenarios employees might face.”

It’s also being used to improve efficiency more generally, in the same way as it might any other business. “The ability to automate high-volume, routine, low-value-added tasks has allowed insurers to speed up their services and increase productivity,” says Steve Bramall, credit director at Allianz Trade. “This frees up valuable experts to spend more time with customers and brokers, improving customer experience.”

The risks and ethics of AI in insurtech

Yet the use of AI also brings risks and ethical considerations for insurers and insurtech firms. “With all AI, you need to understand where the AI models are from and where the data is being trained from and, importantly, whether there is an in-built bias,” says Kevin Gaut, chief technology officer at insurtech INSTANDA. “Proper due diligence on the data is the key, even with your own internal data.”

It’s essential, too, that organisations can explain any decisions that are taken, warns Muylle, and that there is at least some human oversight. “A notable issue is the black-box nature of some AI algorithms that produce results without explanation,” he warns. “To address this, it’s essential to involve humans in the decision-making loop, establish clear AI principles and involve an AI review board or third party. Companies can avoid pitfalls by being transparent with their AI use and co-operating when questioned.”

AI applications themselves also raise the potential for organisations to get caught out in cyber-attacks. “Perpetrators can use generative AI to produce highly believable yet fraudulent insurance claims,” points out Brugger. “They can also use audio synthesis and deepfakes pretending to be someone else. If produced at high-scale, such fraudulent claims can overwhelm the insurer, leading to higher payouts.”

Cyber-attacks can also lead to significant data breaches, which can have serious consequences for insurers. “These can expose confidential client information, which inevitably poses new challenges towards fostering client trust,” says James Harrison, global head of insurance at Dun & Bradstreet. “Additionally, failure to comply with data protection regulations, such as GDPR, can lead to legal consequences and financial penalties.”

Having robust cybersecurity measures is essential, particularly when it comes to sensitive or personal data, says David Dumont, a partner at law firm Hunton Andrews Kurth, and it’s important to ensure these remain able to cope with new regulations. “In the EU, the legal framework on cybersecurity is evolving and becoming more prescriptive,” he explains. “Within the next year, insurtechs may, for example, be required to comply with considerable cybersecurity obligations under the Digital Operational Resilience Act (DORA), depending on the specific type of products and services that they offer.”

AI will augment, rather than replace, human capabilities

All this means AI requires careful handling if insurers and insurtechs are to realise the benefits, without experiencing the downsides. “The future of AI in insurtech is brimming with potential,” believes Colmant. “AI will likely specialise in specific insurance processes, like underwriting or claims management, leading to significant efficiency gains and improved accuracy. This will also likely lead to even greater personalisation and automation.

“However, the focus will likely shift towards a collaborative approach, with AI augmenting human capabilities rather than replacing them entirely. Throughout this evolution, ethical considerations will remain a top priority.”

How artificial intelligence is helping to slash fraud at UK banks – Rob Woods, fraud expert at LexisNexis Risk Solutions, tells Charles Orton-Jones why behavioural data and AI are a powerful fraud-fighting combination

Why is embedded insurance so popular right now? – Charles Orton-Jones asks five industry experts how embedded insurance could transform the sector and whether or not it offers real value for consumers

Will more AI mean more cyberattacks? – An increased use of AI within organisations could spell a rise in cyberattacks, explains Nick Martindale. Here’s what you can do

Avatar photo

Nick Martindale

Nick Martindale is an experienced freelance journalist, editor and copywriter. He specialises in writing about workplace matters, including HR, procurement and technology.