Feature Stories Insurtech Trending

Is the Insurance Sector Risking Over-Reliance on Artificial Intelligence? Part One

This March, The Fintech Times has turned its focus towards insurtech, shedding light on the innovative advancements and sustainable initiatives within the insurance sector.

Today, we explore the role of artificial intelligence in insurance and the delicate balance between leveraging technology and preserving human expertise.

Is the insurance sector risking over-reliance on artificial intelligence, and what’s the balance between innovation and human expertise?

In part one of our spotlight on AI, let’s hear what our community says… (part two here).

LexisNexis Risk Solutions
John Beal, senior vice president, data science, LexisNexis Risk Solutions.
John Beal, senior vice president, data science, LexisNexis Risk Solutions.

Artificial intelligence is already becoming a key area of competition, having a direct impact on the speed of deployment, efficiency, improved customer experience, targeted pricing and customisation,” says John Beal, senior vice president, data science, LexisNexis Risk Solutions. “However, it is not a panacea for all our data problems.”

He continues: “For insurance providers, machine learning cannot effectively manage, cleanse, analyse and deploy data without input from a highly experienced data scientist.

“Algorithms are incredibly helpful and intelligent but without manual intervention by a data scientist, algorithms are not able to accurately structure data or use them for the correct business acumen, especially within Insurance while keeping in mind the relevant regulatory requirements as they model to a loss curve. Understanding the problems, how a solution will work, and how it should be implemented still requires human expertise.

“With human involvement playing such a fundamental role in data and analytics for insurance, ‘applied intelligence’ or ‘machine augmented intelligence’ are better descriptions rather than full artificial intelligence. This is the application of automation within the insurance workflow, alongside the essential human intelligence and business acumen, rather than a fully machine run operational process.

“LexisNexis Risk Solutions has been undertaking data science in this way for more than four decades. While today AI is increasingly helping within the insurance analytics process it is best utilised by a team of expert data scientists who understand the fundamentals of insurance.”

Kevin Gaut, chief technology officer at no-code insurtech platform INSTANDA
Kevin Gaut, chief technology officer, INSTANDA

Kevin Gaut, chief technology officer at no-code insurtech platform INSTANDA, suggests that while artificial intelligence is increasingly ingrained in activities, the idea of replacing humans with machines, especially in fields like insurance, would be too hasty.

‘’AI has seamlessly integrated into our lives, yet the notion of replacing humans with machines, especially in insurance, remains premature. AI, exemplified by Siri or Alexa, operates on predefined rules, executing tasks deterministically.

“In contrast, generative AI, like ChatGPT, uses learning techniques, continuously refining decision-making processes based on available data. These approaches embody different capabilities and characteristics. While ChatGPT has significantly advanced in the last 18 months, the extensive groundwork in machine learning preceding these advancements must be acknowledged.

“Although fears of job displacement are legitimate – Goldman Sachs predicts 300 million jobs worldwide could be affected – the technology is still in its infancy. In the short to medium term, AI’s role should be viewed as a support system, enhancing user capabilities rather than supplanting them. Comparable to a colleague offering assistance, AI provides tools that users can leverage as needed.

“Take underwriting. By streamlining the underwriting process, underwriters can do what they do best – harnessing more data, but quicker and more efficiently. By using a rich vein of data to make better, more informed decisions, AI does not remove human expertise, but saves time and frees underwriters to be even more productive.

“Essentially, AI doesn’t replace human expertise but rather enhances efficiency, allowing individuals to focus on skill development. Looking forward, AI’s trajectory promises continued evolution, presenting opportunities for collaboration and long-term skill refinement.’’

Financial Technology Research Centre (FTRC)
Ian McKenna, Founder, Financial Technology Research Centre (FTRC)
Ian McKenna, founder, Financial Technology Research Centre (FTRC)

Fintech consultancy, the Financial Technology Research Centre is hosting an AI in Financial Advice event this summer. Its founder Ian McKenna says there’s no doubt that AI can offer substantial benefits to the insurance sector.

“While much of the sector is only just becoming acquainted with the potential benefits of AI, human expertise is crucial on several levels. AI providers themselves need to have a very clear and specific use case for the insurance sector to ensure the most effective and valuable outcomes for both the sector and the end consumers.

“Second, insurance providers need to conduct extensive due diligence to ensure that there is a complete understanding of the outcomes generated by using an AI service. This is particularly fundamental following the introduction of Consumer Duty.

“For example, there are cases where it would be unwise to use generative AI due to the risks of hallucinations while predictive AI may be more reliable and auditable. All these nuances therefore need to be assessed on a case-by-case basis with human expertise.”

Stacy Edgar, CEO and founder of Venteur,
Stacy Edgar, CEO and founder of Venteur

It would be a mistake to suggest the insurance sector is over-reliant on artificial intelligence, according to Stacy Edgar, licensed insurance broker and the CEO and founder of health startup Venteur.

“We’re in the early stages of the technology, and while there’s excitement around the tech, the insurance industry is the most risk-averse industry out there. There are valid concerns around data privacy and potential data biases, but these can be overcome if you approach implementation thoughtful.

“In particular, transparency is key. At Venteur, we have publicly share how we trained our AI, where we got the data, and how data is used. This not only helps to meet regulation concerns, but also helps build trust in our AI technology with our clients.”

Carpe Data
Geoff Andrews, chief operating officer at Carpe Data
Geoff Andrews, COO, Carpe Data

Geoff Andrews, chief operating officer at Carpe Data, which provides alternative data to insurance carriers, says the insurance sector is not currently in danger of over-relying on artificial intelligence, as most insurers are still figuring out how and in what areas to apply it.

“Insurance has always had to adapt to changing market conditions and human behaviour, but rarely has it adapted fast. Today the best use cases for advanced generative AI models in insurance are focused on efficiency and accuracy.

“With the looming talent gap presented by an aging workforce, insurers must use AI to maximise the efficiency of time-consuming manual tasks, reduce overall costs (time and resources), and empower human expertise rather than replace it.

“Some examples: AI can automate rating and quoting processes in underwriting so insurers can be more intentional about refining their risk appetite and selection while providing a superior customer experience. And, AI can simplify passthrough toll gates for smaller claims, as well as monitor open claims at scale to flag potentially fraudulent activity.

“Done right, AI will enhance ‘human-in-the-loop’ processes but never entirely replace them. People should be the brain and AI the engine, automating repetitive tasks and organising data-driven insights so people can make more confident decisions with better context while exhibiting the knowledge and compassion integral to insurance.”

Dan Huddart, chief technology officer at specialist home insurer, Homeprotect
Dan Huddart, CTO, Homeprotect

“If anyone’s worried that AI is coming to take our jobs, I think it’s worth looking back at the history of technology in the industry,” says Dan Huddart, chief technology officer at specialist home insurer, Homeprotect. “Just like all technology evolutions, advances in AI will fundamentally change the jobs that we do to meet what customers need and expect from us.

“Insurance changed forever when computers landed on desks. It changed again when the internet linked them all together. Each evolution in statistical modelling has driven a bow wave through the way we calculate prices, analyse risks and design products. Large language models give us new tools and ways to work with text and speech at speeds and granularity that were unthinkable until recently. Advances in image and video technology will drive similar shifts in how we analyse and interact with real world risks and claims.

“Human expertise has been critical through every technology change. For example, the role of an underwriter has the same purpose after each tech revolution, but the tools and productivity per person look very different.

“What AI takes away in single-person productivity, it replaces with new opportunities and new roles. We now gather more data, in new ways, than ever before. Customer expectations go up over time. Risks evolve, and new products need to meet new customer demands. Human expertise is essential in adapting to these new opportunities and we rely on technology including all forms of AI to give us the productivity boost to tackle them.”

Roi Amir is the CEO of insurtech,
Roi Amir is the CEO of insurtech,

Roi Amir is the CEO of insurtech, where he is driving’s mission to work in partnership with insurance companies, building AI and data-led products. He suggests that although it may sometimes feel AI hype is everywhere, it is only pockets of the insurance industry that are adopting AI.

“Incumbents, for example, have still been slow to catch on to the opportunity that technology presents. Key hotspots include marketing, fraud detection, customer service, and claims management and automation. These processes have all existed without AI for many years, but AI can drastically improve the efficiency levels, leading to a knock-on effect across the wider industry.

“Claims AI technology can reduce the time for a claim to be processed from weeks or months, to near real-time, and at a 97 per cent accuracy rate. That being said, some of our research shows that both insurers and customers believe that we shouldn’t totally hand over complete control to AI. By automating the mundane, while maintaining other important elements of the traditional insurance model, we can avoid over reliance. AI should be there to enhance, not replace roles.

Humans are crucial

“There’s been a broad misunderstanding of AI in the insurance industry up until now – that the purpose of AI is to replace claim handlers all together. But the human element is as crucial now as it’s ever been. Our research found that nearly 30 per cent of insurance customers are concerned about losing human interaction where AI is used, and a further 43 per cent lack trust in AI’s decision-making.

“However, different people look for different things in their insurance claims process. In some circumstances, such as a claim on a vet bill, people prefer the journey to be fully automated due to the speed at which AI can solve these cases. Although, people still value human expertise and communication when it comes to more complex scenarios, such as complex medical claims, for which it’s important that the option of that support is reserved. It’s horses for courses, but one should not come without the other.

“Rather than replacing claim handlers, the use of this technology frees up crucial time in the claims process, so that insurance professionals can provide better and a more personal customer experience. With AI innovation, claims handlers will be able to handle more complex claims and spend more contact time with their customers.”

William Perry, VP UK&I and MEA, Medallia
William Perry, VP UK&I and MEA, Medallia

“The insurance sector’s seeming reliance on AI isn’t risky, it’s savvy,” says William Perry, VP UK&I and MEA at management software company Medallia. “Initially fuelled by a need to manage the rising proliferation of data, it has resulted in industry-wide innovation – enhancing the entire customer journey from initial policy purchase, right through to underwriting and making a claim. Such is its perceived value, that Allianz believes it could add $1.1trillion to the insurance market annually.

“AI is arguably just helping the insurance sector to keep pace with the scale of innovation it has to respond to. Indeed, from the rise in autonomous cars with self-driving capabilities, to the continued prevalence of connected devices, insurers need to use tools like AI to connect, use and analyse the data generated – or risk being left behind.

“As with all successful implementations though, technology must be coupled with the requisite human expertise if it is to reach its potential. Securing buy-in from the C-Suite to invest in engaging the right people to mould the strategies to harness the full power of AI, will be essential over the coming months and years.”

Will Larcombe, co-founder and director of Stellarmann
Will Larcombe, co-founder and director of Stellarmann

Will Larcombe co-founded technology and change delivery consultancy Stellarmann in 2020, with his business partner Alex Colwell. He thinks investing in AI is a must.

“The greatest risk to the insurance sector is surely not investing in AI, he says. “Businesses cannot afford not to, if they are to meet customer expectations, drive time and costs saving efficiencies, gain competitive edge, and detect fraudsters – who are themselves a step ahead with GenAI.

“There are literally hundreds of potential applications for AI in insurance though, so businesses must prioritise by identifying the solutions that make most sense for their specific operations, customer base and architectural framework.

“Highly specialist human expertise is essential to successfully manage the introduction of AI and businesses will need to bring in expertise in the following areas:

  • Understanding the businesses areas where AI will bring most value
  • Implementing the technology
  • Maintaining up to date, relevant, clean data – as AI is data dependent
  • Negotiating changing regulations
  • Ensuring all stakeholders know what and why things are changing

“The talent pool in this emerging area is still small and relatively undefined. As such, finding and and retaining these new skills and experience is the first step towards successful AI adoption and so competition for human expertise will be high. To attract the best and get the most from AI implementations, organisations will need to show they are serious about investing in AI.”


Related posts

Yolt Finds Open Banking Could Save SME Online Retailres over £19,000 a Month in Transaction Fees

Polly Jean Harrison

FCA’s push to enforce transparency as “welcome news” for payments sector.

Manisha Patel

UK Fintech Curve Fastest Startup Ever to Reach £4m Crowdfunding on Crowdcube

Mark Walker