Insurers are rapidly responding to the increasing adoption of AI by either excluding AI-related losses entirely from corporate liability policies or by introducing sublimits that cap payouts for AI-stemming risks. This trend is driven by the unpredictable nature of AI agents and the potential for systemic risks across the economy, making AI damage difficult to insure. Kevin Kalinich, AI risk transfer leader at Aon, highlights that developers often "don't know what it's going to do 100 per cent of the time," making it fundamentally different to insure compared to traditional perils like hurricanes or terrorist attacks.
This shift by insurers is creating substantial gaps in coverage for policyholders, who may be exposed to significant liabilities from AI-related mistakes. Lawyers and insurance brokers anticipate that it will take years for courts to determine liability across AI developers, businesses deploying the technology, and end-users. The situation is being closely watched, particularly in cases like Derek Mobley's lawsuit against Workday, where he claims discrimination by AI algorithms; a court has allowed the case to proceed, signaling potential far-reaching consequences beyond employment discrimination.
Some analysts, like Gretchen Hoff Varner at Covington, view this as an opportunity for insurers to develop new products, likening it to how cyber insurance emerged as a standalone offering after exclusions for tech-related losses. However, the magnitude of potential AI-related losses, which could run into hundreds of millions or even billions of dollars, is a major concern. The uptake of AI also introduces systemic risks, especially for critical infrastructure like water and electricity, which could lead to widespread disruption, further complicating insurance coverage. An example of a recent large-scale verdict is the case against Meta and Google, where a jury found them negligent for social media platforms designed to be addictive to children, demonstrating the potential for significant legal and financial consequences related to technology. This suggests that AI losses could stretch even further than current tech liabilities, highlighting the urgency for clear insurance frameworks.
