How accurate data and AI can transform claims and help customers build resilience

Artificial intelligence (AI) is helping transform insurance, including claims. AI presents an exciting opportunity to enhance commercial claims, from improved customer interactions and service levels to helping reveal insights and informing loss prevention actions.

Raising service levels

An obvious application of AI in claims is to support customers in their hour of need. Generative AI enables more intuitive chat bot tools to address customer queries 24/7, speed up processes and response times, and generally improve communication. AI enables insurers quickly assess claims and put them on the right – such as directing a claim to the appropriate claims team or automatically instructing a loss adjuster.

A fast and more efficient response is particularly important for large complex commercial claims – especially those with the potential for significant business interruption - where the best outcome for both parties comes from early intervention. Zurich has been using AI to help predict likely outcomes from claims notifications in a bid to identify those claims that are likely to escalate, and therefore customers benefit from early intervention and loss mitigation.

Freeing up resources

AI can also make claims handling more efficient, reducing the time claims managers and loss adjusters spend on administration, and freeing up resources to focus on value-add activities. For example, AI can automatically read and summarise claims reports and correspondence, and then propose or take appropriate action. A process that currently takes a claims manager hours or days can be completed in seconds. McKinsey predicts that more than half of claims activities can be automated by 2030, while research from Oliver Wyman suggests that automating tasks could achieve up to 20% time saving benefits.

In addition to more efficient processes, AI also enables insurers to better monitor the cost of claims, identifying claims leakages and fraudulent claims, as well as through a better understanding of claims inflation and cost drivers. AI-powered claims insights can also feedback to underwriting, risk engineering and consulting services, which leads to better understanding of risk, and allows insurers to serve their customers better. In the end it is about supporting customers by understanding the true impact of claims.

Streamlining catastrophe claims

AI has already helped Zurich make significant improvement in catastrophe claims. Last year Zurich piloted Catastrophe Intelligent Agent (CATIA), an in-house AI-powered tool to streamline the claims tagging process. In just minutes, it identifies catastrophe claims based on the cause of loss and claim descriptions, targeting more accurate reinsurance recoveries. “The CATIA tool allowed us to combine traditional AI techniques together with new Generative AI capabilities, and to demonstrate its effectiveness”, says Christian Westermann, Group Head of AI of Zurich Insurance Company.

By automating the identification and validation of catastrophe claims, CATIA has reduced the time spent on these tasks, removing manual review, and improving operational efficiency. During its pilot phase last year, the tool uncovered an additional 500 catastrophe claims for five natural catastrophe events across five countries, generating savings of $1.4m, easily recouping its initial investment. And with faster and more accurate tagging of catastrophe claims, the claims experts can shift the focus from admin to focus on customers' needs and providing support, for example by assessing the impact of the damage and securing local resources on the site of the catastrophe earlier than previously would have been possible.

CATIA is a good example of how simple cost-effective AI solutions can make a big difference and help increase efficiency. Building on its success, Zurich sees exciting opportunities to use AI for further advancements in cause of loss determination and mining for untapped claims insights that were previously difficult to reach.

Powering loss prevention

With advancements in technology, the application of AI in insurance has moved to another level. AI capabilities are now more powerful, in particular due to the introduction of GenAI. AI enables insurers to explore what might happen next, and suggest potential actions. Such capabilities will present new opportunities to use claims insights to pro-actively prevent and mitigate losses, as well as enable more innovative services.

Zurich is already using AI to identify insured properties most at risk of fire and water damage claims, enabling claims, underwriting and risk engineering to work with customers to prevent losses. In the UK, Zurich deployed AI to gain a deeper understanding of the risk of fire based on various indicators such as proximity to a fire hydrant, response time from the fire brigade and other relevant data points. This is just one example of how insurers can leverage external data and data from previous events to support customers in building resilience.

Insurers need to leverage claims data and AI to effectively address the increasing risks associated with climate change. The combination of geospatial data, images and claims data can unlock powerful insight to manage our portfolios and enable better decision making.

Quality, accurate data is key to unlocking AI

AI is a reminder of the fundamental importance of good quality data for claims handling, as well for underwriting and risk engineering. Not only do insurers need accurate information, but they also need to accurately record it. Incorrect or incomplete data is as good as no data at all, while analytics and AI can bring little value without accurate data.

AI has the potential to unlock unstructured data, resulting in more granular claims data, which should in turn produce better insights into loss trends, drivers for claims costs, and issues like under insurance and the effects of inflation.

Customers can, in turn, benefit from more targeted risk management. With this data, insurers are able to provide more insights to customers based on loss trends in their industries or regions and support customers in taking measures to prevent future losses. Ultimately, all of this helps organizations become more resilient.

“Thanks to its language capabilities, Generative AI offers new possibilities to understand claims and larger losses at a much more granular level, no matter in which language the claim was originally recorded”, says Westermann. “As such, the understanding of a claim and its embedding in a wider perspective does not stop at the country border.”

Traditionally claims data has been collected by insurers for financial and accounting, as well as to manage resources. However, AI is unlocking the power of more granular claims data to understand risks and the effectiveness of loss mitigation actions.

As an industry we must strive to improve the quality and completeness of data to better serve the needs of customers and partners. Robust data helps an insurer handle a claim faster and more efficiently, as well as support customers better with products and services.

Mutual benefits

Accurate data collection, coupled with AI, can help insurers better understand customer needs and provide a tailored service. By collecting more granular claims data, insurers can support better risk management and craft more effective policies. This data can be used to predict, prevent, and address risks, ultimately benefiting the customer.

Originally published in Commercial Risk on July 2, 2024.