Insurance Model Risk: Audit | Moring
An insurance model risk audit is where the AIS Program meets the examiner: the AI system inventory, the written governance program, and risk evaluations proportionate to impact. The inventory is usually the first question, and vendor-embedded AI is often the largest unrecorded part of the inventory.
Short answer
An insurance model risk audit is where the AIS Program meets the examiner: the AI system inventory, the written governance program, and risk evaluations proportionate to impact. The inventory is usually the first question, and vendor-embedded AI is often the largest unrecorded part of the inventory.
How it works
The audit is where the AIS Program is proven. The inventory is usually the first question in an examination, and the vendor-embedded AI is often the largest unrecorded part of the inventory. A first pass is a matter of days rather than weeks when the queries are right: IAM identities that belong to non-human actors, procurement records for AI-flagged vendors, and workflow owners naming the systems that touch their processes. The effort grows with the number of business units, vendors, and adopting states involved, and the inventory is the control that scopes every other control.
The written program is the document that answers who governs. It should be written around the inventory, not in the abstract: which systems are high-impact, who owns each, how they are evaluated, how exceptions are handled, and how the program changes as systems change. A program written before the inventory describes an organisation that does not exist, and the mismatch is visible to anyone who compares the document with the systems.
The risk evaluations are proportionate to impact. Each system gets an evaluation proportionate to what it does, and the classification is by use case, not model: the same model is low-risk summarizing meeting notes and high-risk screening job applicants. For an insurance carrier, the classification is by the workflow the model touches: underwriting, pricing, claims administration, fraud detection, and policy servicing. The highest regulatory scrutiny in the stack is underwriting and risk selection.
The audit is the control that makes the NAIC AIS Program enforceable.
The NAIC Model Bulletin, adopted by roughly 24 states, requires insurers to adopt, implement, and maintain a documented AIS Program, and the bulletin expects records sufficient to show the program operates and to produce to regulators on request. The EU AI Act places human oversight at the center of high-risk systems, including the ability to intervene in or stop operation when needed.
Frequently asked questions
What does an insurance model risk audit examine?
The audit examines the AI system inventory, the written governance program, and the risk evaluations proportionate to impact. The inventory is usually the first question in an examination, and the vendor-embedded AI is often the largest unrecorded part of the inventory. A first pass is a matter of days rather than weeks when the queries are right: IAM identities that belong to non-human actors, procurement records for AI-flagged vendors, and workflow owners naming the systems that touch their processes.
How should an insurance carrier write the AIS Program around the inventory?
The written program should be written around the inventory, not in the abstract: which systems are high-impact, who owns each, how they are evaluated, how exceptions are handled, and how the program changes as systems change. A program written before the inventory describes an organisation that does not exist, and the mismatch is visible to anyone who compares the document with the systems. The inventory is the control that scopes every other control.
How should an insurance carrier classify model risk for audit purposes?
The classification is by use case, not model: the same model is low-risk summarizing meeting notes and high-risk screening job applicants. For an insurance carrier, the classification is by the workflow the model touches: underwriting, pricing, claims administration, fraud detection, and policy servicing. The highest regulatory scrutiny in the stack is underwriting and risk selection, and the risk classification is the control that makes the oversight proportionate to the risk.
What does the NAIC Model Bulletin expect an insurance carrier to produce in an audit?
The NAIC Model Bulletin, adopted by roughly 24 states, requires insurers to adopt, implement, and maintain a documented AIS Program, and the bulletin expects records sufficient to show the program operates and to produce to regulators on request. The audit is the control that makes the NAIC AIS Program enforceable, and the EU AI Act places human oversight at the center of high-risk systems, including the ability to intervene in or stop operation when needed.