NeuralMetrics technical co-founder Marcus Daley examines the ethical dimensions of employing AI in commercial underwriting. Learn how AI can improve underwriting while complying with regulations and avoiding ethical pitfalls.
In a competitive commercial insurance market, quick risk assessment, quotes, and policy binding are vital for insurers. However, the evolving regulatory landscape requires insurers to be mindful of their data sources. FTI Consulting has recently identified potential issues with some data sources. NeuralMetrics technical co-founder Marcus Daley discusses how AI is revolutionizing commercial underwriting and provides guidance on navigating ethical considerations.
Q: What opportunity does NeuralMetrics see in commercial insurance?
NeuralMetrics recognizes a significant opportunity in commercial insurance, particularly for small and medium businesses (SMBs). This market segment is often underserved, making it challenging for SMBs to obtain insurance coverage. Business owners view insurance as an expense and lack familiarity with the intricacies of the industry.
From the carrier’s perspective, NeuralMetrics seeks to transform this high-volume market into a high-margin opportunity. By creating an open, unbiased data solution for risk assessment that facilitates precise coverages and fair premium pricing, carriers can more effectively expand into the SMB market. This approach enables businesses that previously struggled to secure adequate insurance to access coverage more efficiently.
On a broader scale, AI-powered classification and risk assessment data solutions have the potential to reshape the culture of insurance transactions in the SMB market. The goal is for business owners to view insurance as a strategic tool for de-risking and positioning their enterprises for growth, rather than just an expense.
Q: While AI and data improve the underwriting process, how can insurers ensure ethical use and regulatory compliance?
When evaluating data about companies, it’s crucial to avoid issues that may raise ethical concerns. For example, insurers should steer clear of demographic data, as making assumptions based on such information can be ambiguous and inadequate.
Personal identifiable information, such as information about business owners, management, and the board of directors, should not be sourced from the internet using AI. Factoring in individual information during risk assessment can compromise privacy.
Insurers should also be aware of potential risks associated with AI. For example, image recognition can be a powerful tool, but it can introduce bias if not programmed carefully. Image recognition AI learns from human interpretations and associations of value, potentially inheriting biases from human “teachers.” To mitigate bias, it’s important to assess whether the information can be applied across industries and linked back to a source document, ensuring valuable data insights for risk assessment without introducing bias.