As AI systems move from experimentation into mission-critical business processes, a new category of risk is emerging that traditional insurance frameworks were never designed to address. This session explores where today’s policies fall short when it comes to AI-specific exposures and how the market can begin to distinguish between uninsurable AI hype and genuinely underwriteable AI performance risk.

Drawing on real underwriting and product development experience, the session will examine how clearly defined AI failure modes, measurable performance thresholds, and parametric-style triggers can transform abstract AI risk into something insurers can price, structure, and scale. The discussion will also cover the implications for brokers, carriers, and reinsurers as AI liability shifts from “if something breaks” to “when outputs are simply wrong.”