Insurers urged to turn risk insight thinking on its head to profit from AI

Insurers need to take a different approach to the way they look at underwriting to get the most out of new technology, according to Laila Beane, chief customer officer and head of consulting at Intellect SEEC.

Beane said: “Given that insurers have been very linear in their way of thinking, there's a lot of manual touchpoints and therefore there are a lot of costs upfront for every broker submission. “There are not many risk insights except from the underwriters, who have some traditional sources that are paid [for] which they get information from. Underwriters also use Google search and have historical insights.”

She explained that her company wants to flip that thinking on its head, as she gave a presentation, titled ‘An AI-Powered, Data-First Approach To A Quicker, Better And More Profitable Portfolio’, at Intelligent Insurer’s Underwriting Innovation USA virtual event.

Beane said that the key question to consider was how can you drive value to every business knowing that every carrier is unique as every carrier is slightly different, with different lines of business?

Although 100 percent of submissions generally hit the underwriter queue in most carrier environments, there is what she called a certain amount of ‘broker submission pack triage’, inclusive of judgement calls and other forms of reviews occurring by several linear manual touch points even before it reaches the underwriter.

But by using artificial intelligence (AI) in the form of proven insurance-trained machine-learning models companies can reduce submission triage time drastically; improve data quality; enrich the risk data footprint for the account; perform appetite matching up front; and ensure higher bind rates in desired portfolios, she said. This kind of technology can also reduce expenses and improve accuracy, she added.

In her presentation, Beane went through a number of scenarios in which she emphasised the advantages of increased use of AI in analysing the bundles of insurance business that get submitted every day across the market. She also highlighted the value of using AI to analyse key risk elements from the widest possible range of data sources.

She acknowledged that some in the market might not be comfortable with increasing the use of AI.

“It's a cultural shift in thinking,” said Beane. “The biggest challenge with all these AI projects is not really the technology, the challenge is more of shifting gears and change management.”

Despite some industry reluctance to embrace AI, she said one way to view the advantages of this technology is that it can offer a clearer portfolio overview. AI can allow companies that are struggling to look at their current portfolios to gain a better grasp of risks within it, which will give them a better idea of what is a good risk these days and what is not. And AI can provide even more data by using the internet to search additional unstructured information that underwriters just might not have time to look for, such as social media, about the companies that are being insured or which are connected to a risk.

Asked for some examples of this last point by the audience, Beane said that analysing employee and customer sentiments can provide deep insights into accidents and injuries in the workplace that might be relevant to a company’s operational health, employee training and safety protocols.

Finally, looking forward to 2021, she said that Intellect is looking at a number of areas with its partners, including the streaming data (via the internet of things) and expansion of proactive loss prevention alerts.

To watch the full session  click here.