Insurance
Home›Insurance›Industry & Deals›AI agency models raise new insurance questions for del…
AI agency models raise new insurance questions for delegated computer systems
The framework at issue, Agent Insurable Value, seeks to separate the economic exposure created by AI authority from the cost of models or compute.
NVIDIA has argued that AI-factory compute is becoming an investable infrastructure asset, supported by financing platforms including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR that are designed to mobilize more than $500 billion in third-party capital over time, Insurance Journal reports.
The commentary says the underwriting logic for that kind of infrastructure focuses on factors such as customer quality, demand, utilization, cash flow, and residual value, reflecting a shift in how artificial intelligence is viewed, from a software expense to productive infrastructure that turns energy and data into intelligence.
Insurance, however, is presented as needing to address a different problem once AI systems are authorized to act, because the system’s economic significance would not be captured by the model cost or compute used to run it.
To quantify that exposure, the piece describes Agent Insurable Value as a framework that estimates the economic risk created when an organization delegates meaningful authority to an AI agent, distinguishing it from the software’s market value or the infrastructure replacement cost and noting that even an inexpensive agent could influence major business activity.