Speaking on a panel at the Insurtech Insights Conference in London, our CEO Richard Hartley discussed how AI is impacting insurance pricing and enabling underwriters.
The panel, titled “Dynamic Pricing & Long Term Underwriting using AI”, was moderated by insurtech expert and co-founder of Instech London, Matthew Grant. Other panelists included Laurent Rousseau, Deputy CEO of SCOR Global P&C, Marcus Winter, Head of Reinsurance Development at Munich Re, Michael Natusch Global Head of AI at Prudential, and Ryan Kottenstette CEO and Co-founder of Cape Analytics.
Below we recap some of the key takeaways.
The insurance industry is at an inflection point, and it’s an exciting time to be involved
The panel agreed that insurance is starting to undergo the same transformation as finance did a decade ago.
“You can only really appreciate inflection points in an industry if you’ve seen them in others,” commented Laurent Rousseau, who began working in insurance after a career in investment banking.
Compared to insurance, risk-taking in finance has typically been more strategic and data-driven. Big data and AI have already revolutionised financial trading, and algorithms are used market-wide to analyse reams of data and power accurate and efficient decision-making.
In insurance, people are starting to realise benefits of using internal and external data to build a better understanding of risks, data collection is now becoming more structured and standardised to support sophisticated analysis, and customer expectations are forcing improvements in efficiency across the entire insurance value chain.
The role of the underwriter is changing, but it won’t be replaced by AI
Although AI is increasingly becoming embedded in the underwriting process, it is an enabler rather than a threat, as it can help underwriters to make faster, better-informed decisions, and liberate them from repetitive low-value tasks. Underwriters will remain essential to the industry.
Large, complex risks will continue to be assessed by human underwriters for the foreseeable future as AI requires large volumes of data about similar classes of risk in order to begin making accurate predictions.
SME insurance is a great candidate for automated underwriting as these types of risks have similar characteristics, and the necessary data to build performant risk selection and pricing models exist today.
The panel agreed that for smaller classes of risk such as SME, automation is also necessary as the premium is so low.
“Our customers tell us that for smaller classes of risk, as soon as an underwriter has to manually assess a submission, they’ve lost money,” said Richard Hartley.
However, underwriters will remain an important part of the process. “In finance, you need a seasoned trader to override the algorithms when things go wrong. The same thing applies to insurance,” said Laurent Rousseau.
On partnering with insurtechs
The panel agreed that partnerships between insurers and insurtechs are fundamental in driving the industry forward.
By partnering with traditional, trusted insurers, Insurtechs can test and refine their products rapidly, gain access to a wider consumer base, and make use of decades of insurance expertise.
Through partnering with insurtechs, incumbents gain access to talented entrepreneurs and technologists who are building solutions which can get insurers closer to their customers and help them to make their conventional workflows better, faster, and more cost-effective.