DGTAL: Gen AI can identify emerging risks
Technology
Readingtime: 5 min
**Generative artificial intelligence (Gen-AI) can be a powerful tool for identifying and pricing emerging risks, according to one of the founders of a pioneering insurtech. **
Vanda Giannara, the co-founder and chief client officer of AI insurtech DGTAL, told Monte Carlo Today that Gen-AI could be used to capture data about emerging risks in a way that is impossible for the re/insurance industry to do so far.
This would enable the industry to segment, classify and price the risks much more quickly and accurately, she said.
DGTAL is using its product, Driller, to support pricing actuaries and underwriters in identifying and capturing qualitative information in claims that have no historical data, such as emerging risks and trends.
Driller was launched last September and is currently being used by re/insurers for AI audits and as a claims assessment co-pilot. The AI-powered portfolio tool is designed to develop and display a contextual understanding across a wide array of cases and documents. It allows users to sift through massive volumes of data to identify patterns, trends, and insights.
“Driller is able to analyse packages of up to 500 pages of claims documents in 35 minutes.”
It is particularly powerful at drilling into unstructured documentation which could take weeks or months to be assessed manually, says DGTAL. Instead, it can do it in minutes.
That application is being used by eight companies and another four are carrying out proof of concept testing with the technology.
DGTAL made headlines earlier this year when French re/insurance giant SCOR said it was carrying out a pilot with Driller to assess disability claims. Giannara said SCOR had recently accepted the proof of concept.
“Driller is able to analyse packages of up to 500 pages of claims documents in 35 minutes. It would take a person a week or more to review the same claims packages,” she said.
The company is now looking at how Gen-AI can be used to identify emerging risks.
“Gen-AI is a very competent tool for language understanding, for reading files, processing and giving you this information,” Giannara said. “There are three reasons Gen-AI can be used in the emerging risk space that is becoming very important.
“The first is the speed at which things change. We have always had emerging risks in insurance, but right now, by the time you realise something is an emerging risk or trend, it is almost mainstream. I call this ‘the new normal’. The present is the past and the future is happening as we speak.
“This significantly affects the development of a claim,” she said. “You have to have a more powerful lens in order to capture this data.
“With an emerging risk there is not the historical data around it. If you take the example of long COVID, it can be the underlying issue for pneumonia and other respiratory illnesses in the last 12 months,” she said.
“You cannot distinguish, unless you read a claims file, whether the underlying cause was long COVID or if there was another cause. Someone needs to read all the claims files to get this understanding but that does not happen, so a pricing actuary cannot deep dive and apply his pricing model.
“AI is able to detect within the last 12 months—from the fresh claims data—the proportion of people who had pneumonia with long COVID as the underlying cause and what the average claims payment was.
“You can have a very good actuarial approach to an emerging risk which is not only probabilistic and mathematical but is complemented with aggregated fresh claims data that cannot be captured anywhere now.
“Then you can apply all your actuarial modelling to it,” she said. “You are able to narrow down very nuanced and sophisticated cohorts which is great for pricing actuaries and risk managers.”
Changing behaviour
Another way that Gen-AI could benefit the health insurance industry is through the use of wearable technologies such as Fitbits and other health trackers, which could be used the same way as telematics are used to incentivise the behaviour of drivers.
Giannara said if people with wearables were more conscious of healthy behaviours, insurers would be able to reward insureds for good behaviour by reducing premiums.
“Technology has the ability to incentivise or disincentivise behaviour and create great new products based on understanding of risk,” she said.
She agreed that such a programme could be devised without Gen-AI, but where it becomes useful is to link claims information with underwriting data in one single source of truth, always with the policyholder’s consent, in order to take a personalised and customer-centric approach to product development.
“The tech exists today. Good data need effort and we need to focus more there. Find narrow solid use-cases and keep trying. The results can be astonishing,” she concluded.
Vanda Giannara is the co-founder and chief client officer of DGTAL. She can be contacted at: v.giannara@dgtal.io
For more news from the Rendez-Vous de Septembre (RVS) click here.