Transforming Insurance Claims with Artificial Intelligence (Claims Journal)

Transforming Insurance Claims with Artificial Intelligence

Tuesday, June 4th, 2024 Fraud Insurance Industry Risk Management

At Carrier Management’s annual InsurTech Summit, Frank Giaoui of Optimalex Solutions and Shane Riedman of Verisk explored how artificial intelligence (AI) is revolutionizing property/casualty insurance claim processing. Their insights highlighted the technology’s potential to enhance efficiency, detect fraud, and improve fairness in claims settlements.

Benefits and Opportunities of AI in Claims Processing


Accelerating Human Tasks:
Shane Riedman emphasized that current AI technology enhances rather than replaces human reasoning and judgment in claims processing. Large language models, for instance, can address workforce challenges by maintaining the knowledge and skill sets essential for frontline insurance operations.

Reducing Loss Ratios:
Frank Giaoui identified claims management as a significant opportunity for insurers to reduce their loss ratios. His company, Optimalex Solutions, uses AI to predict fair, timely, and consistent settlement values. AI helps avoid the randomness and inefficiency in compensating damages, ensuring fairer outcomes for all parties involved.

Fraud Detection:
Riedman highlighted Verisk’s use of AI to combat claims fraud. As fraudsters employ advanced image manipulation techniques, Verisk develops detection mechanisms using similar AI technologies. Their machine learning models, combined with large language models, significantly speed up the fraud detection process, reducing it from days or weeks to mere hours.

Combating Nuclear Verdicts and Social Inflation


Data Sharing and Federated Learning:
Giaoui pointed out that while the plaintiffs’ bar embraces AI and data sharing, insurance carriers have been hesitant. He advocates for anonymized, secure, and compliant data sharing within the industry to access federated learning. Optimalex leverages external data through learning models to make accurate predictions without compromising data security.

Workflow Automation:
AI optimizes time management by automating data collection and structuring processes. According to Giaoui, up to 60% of claim adjusters’ time is spent on documentation rather than resolution. Automating these tasks allows claims experts to focus on resolving claims, where their expertise is indispensable.

Addressing AI Bias


Rigorous Testing for Bias:
Riedman stressed the importance of testing AI models for bias. Verisk’s approach includes avoiding data inputs that could serve as proxies for protected statuses, employing peer reviews by experienced data scientists, and using technology specifically designed to detect model bias. This multi-pronged strategy ensures fairness and accuracy in AI-driven claims processing.


External References & Further Reading
https://www.claimsjournal.com/news/national/2024/06/04/323924.htm
SOS Ladder AssistMid-America Catastrophe ServicesAspen Claims ServiceNationwide OversprayHancock Claims Consultants