The experimental system builds on Marsh’s Risk Companion platform to simulate how business decisions affect risk exposure.
Marsh is testing an experimental AI tool, “Project Leapfrog,” with a select group of its insurance clients, designed to highlight the business and insurance exposures associated with corporate decisions in real time, according to a recent interview with Marsh’s John Davies at Insurance Business Magazine.
According to Marsh, Project Leapfrog combines large language models (LLMs) with proprietary risk datasets and loss-modeling tools to simulate the downstream effects of business decisions. These may include acquisitions, operational changes, or technology deployments, and how those actions alter a company’s exposure across areas such as cyber risk, liability, and insurance exposure.
The system is an outgrowth of Marsh’s recently launched AI-powered Risk Companion platform. Risk Companion employs Marsh’s large risk data repository and actuarial models to report on client risks and provide mitigation solutions at speed. Leapfrog goes further by analyzing high-level business decisions to gauge the potential risks and exposures associated with them.
For example, Leapfrog can model how acquiring a company with different cybersecurity controls could affect the buyer’s insurance profile. The system generates multiple scenarios in real time, allowing risk managers to evaluate trade-offs before finalizing decisions.
The recent launch of Risk Companion and now Leapfrog highlights a shift in Insurance toward continuous, model-driven risk assessment, particularly as insureds incorporate AI into decision-making processes that have direct financial and insurance implications.
