This category gathers foundational reference articles on the insurability of AI systems in production. It defines how generative and agentic systems enter insurance contracts, how evidence can be produced over time, and where coverage stops.
The series distinguishes Silent AI from dedicated insurance products, static audit from continuous behavioral evidence, and measurement signals from insurance verdicts. It is complementary to the AI Governance Reference corpus: metrology documents observable behavior; this series examines what that documentation can, and cannot, support in underwriting, claims, and cover design. NeoMundi measures. It does not insure, price, or decide liability.
Introduction Public discussion of AI risk is still skewed toward a small number of visible cases. Those cases matter. They do not describe the loss that is forming inside ordinary production workflows: repeated, mid-severity incidents that erode margin, delay operations, or generate handling costs before anyone treats them as a claim. For insurance, the relevant […]
Introduction Enterprise underwriting still relies, in large part, on a file assembled at a point in time: questionnaires, architecture descriptions, test results, and a compliance dossier. That method is coherent for a deterministic system whose executable behavior is expected to remain the same between two versions. It is a weak method for a generative or
Introduction Insurance can transfer a risk only when that risk can be described, bounded, and evidenced. Many AI systems in production fail this test. They change after the contract is signed. Their failures are often distributed across a model provider, an integrator, a prompt, a knowledge base, and a human workflow. The resulting loss may
Introduction For several years, losses linked to AI systems have been absorbed by enterprise policies that were neither designed nor priced for that exposure. The mechanism is not new. It reproduces, in another technological cycle, what the market already observed with Silent Cyber. Silent AI names that situation: a generative or agentic risk that sits
