Using AI tools to mitigate climate change insurance risks

This article was produced in partnership with Crawford & Company

Desmond Devoy, of Insurance Business, sat down with Sam Krishnamurthy, chief technology officer, digital solutions, at Crawford & Company, to discuss how AI can both help predict environmental catastrophes like forest fires, and streamline the insurance process after such an event.

Climate change continues to be a risk on the minds of insurers, brokers and risk managers worldwide – but can AI be a tool in mitigating it?

Sam Krishnamurthy (pictured), Crawford & Company’s chief technology officer, digital solutions, believes so.

Climate change, and the risks associated with it, have been a concern for many years. The latest United Nations Climate Change Conference, held in Dubai this past November, saw participating countries agree to begin to ‘transition away’ from fossil fuels. The threats of climate change were clearly evident during discussions – from soaring heat in India, to forest fires in Canada, and flooding in America. In the UK too, a recent surge of storms, culminating most recently with Storms Isha and Jocelyn, which caused an array of flood damage, have been linked to climate change.

“We see a lot of evolving risks linked to climate change,” said Krishnamurthy, during a recent interview.

He pointed to continued deforestation, greenhouse gasses, and other pollutants, which are creating a multi-faceted threat impacting the world’s ecosystem. Still, there is a ray of sunlight amid the dark skies.

“I think AI can be helpful to mitigate some of these risks,” he said. “There are a number of areas where we can leverage AI.”

Harnessing AI to help with environmental events, before and after

Many areas of the insurance ecosystem are already creating their own AI models, which can predict climate patterns and extreme events, so that better, more timely preparations and responses, can be made.

Some companies are already using AI, for example, to predict patterns for forest fire events, by using sensors out in the forest, and to alert policyholders in the vicinity of a possible fire risk.

“Many parametric insurance companies are leveraging large-scale, diverse data sources, such as IoT, satellite, and sensor data, using advanced algorithms for machine learning to enhance underwriting insights. They are providing sensors for the policyholder free of cost,” he said. “For example – a smart sensor which monitors a home’s electrical network for early signs of fire risks like micro-arcs and sparks, enabling early intervention to prevent potential fires. This kind of proactive risk mitigation eliminates potential threats that could lead to significant losses in the future.”

Parametric insurance options are growing, with tools installed in businesses or residential homes that can automatically alert the policyholder on what is happening as soon as a catastrophic event occurs. These tools can automatically trigger a payout too, versus a more traditional wait time of between 60 and 90 days. Despite its benefits, the adoption of parametric insurance is still limited compared to traditional indemnity insurance, necessitating a focus on accurate data analytics and staying abreast of evolving regulatory requirements.

But what is Crawford, specifically, doing with AI?

“At Crawford, we are leveraging AI across most of our businesses to uplift operational, customer, and employee experiences with a responsible AI framework,” Krishnamurthy said. “This integration of AI with human expertise in claims management ensures efficient processing and effective decision-making.  Also, we’re particularly focused on the role of AI in claims automation. After a major event like a hurricane, carriers typically face thousands of claims. Crawford is harnessing AI in its product innovation space, working to transform the claims process.

“As an example, starting with the FNOL (first notice of loss) stage, AI aids in efficiently processing large claim volumes and providing accurate coverage reviews, aiding adjusters in making timely, effective coverage decisions. It also streamlines policyholder interactions with assessing damage pictures for fraud detection and intelligent triage assists adjusters in optimally routing and managing inspections, thereby accelerating claim resolutions, and improving better outcomes for our policyholders and clients.

“However, one of the big concerns is AI decision-making models may develop inherent biases, stemming from the biased data used in their training. Hence, we emphasize a thorough data privacy review when it comes to review training datasets and also the ‘human in the loop’ approach is essential to review AI’s outcome. At Crawford, the Digital Desk platform manages digital claims with desk adjusters overseeing AI-directed claim triage and channel segmentations. Adjusters review AI decisions using confidence scores: high scores expedite routing claims to the right channels, while low confident triage core might prompt model retraining from adjuster feedback. This not only ensures precise and unbiased claim routing, but also builds trust to drive better outcomes.”

With an increase in claim complexity challenges, organisations need to streamline their coverage review processes. Crawford’s GenAI technology solution rapidly assesses claim coverage against document evidence and cross-checks it with policy terms. This aids desk adjusters in making informed decisions, automates coverage review and streamlines the entire process, cutting down on administrative overheads.

AI as a policy enforcement tool

But there’s more that AI can do.

“AI can be used to support the enforcement of environmental policies, including monitoring, compliance, and reporting of any kind of violations,” explained Krishnamurthy. “We are developing solutions where our adjusters, or claim handlers can utilise an AI platform similar to Chat-GPT to gain insights about claim coverage against a policy while they assess coverage and investigate claims. This means information can be gleaned in a matter of minutes or seconds, determining whether coverage is offered within a policy or not. It also enables them to respond to policyholders much sooner, thereby enhancing the customer experience. This coverage review process improves the accuracy and auditability of claim reviews, directly extracting insights from policy documents. Moreover, it strengthens the financial health of insurers by minimising claim errors and optimising reserve allocation, accelerating claim processing, and ensuring regulatory compliance.”

Crawford’s future commitment to AI

2024 is seeing more digital solutions come to the fore, as Krishnamurthy is partnering with  Crawford’s business stakeholder, Kenneth Tolson, global president, digital solutions,  in heading the digital product ecosystems with a focus on re-imagining and simplifying the claims process and improving customer experience outcomes.

“Our GenAI product is in the MVP stage right now,” he said. “We are going to roll out these products in the near future.”

He is hoping for wider awareness of these AI tools among insurers and brokers. He pointed out that Crawford has a large portfolio of clients, both carriers and brokers, that the company is trying to educate about its expanding AI framework.

“Insurers and brokers can benefit from increased operational efficiency and reduced costs due to automation,” he said. “This human in the loop approach to review AI’s decisions enhances consistent claim handling for effective risk management and provides a competitive edge in the market through the adoption of advanced technology. Most importantly, we always strive to keep customer consent, data privacy and security on the forefront so that we can build a responsible AI framework.”