Loss Scenarios
Loss Scenario Builder
Build an explicit path from AI risk to event, trigger, gross loss, controls, modeled loss, retained exposure, and potential coverage pathway.
Runaway AI Resource Consumption
An autonomous agent recursively invokes high-cost models and external APIs.
Gross loss
$4,200,000
Before controls
Modeled loss
$2,620,800
After control adjustment
Expected loss
$419,328
Frequency-weighted scenario
MPL
$5,040,000
Maximum plausible loss
Scenario pathway
- Risk: ARAR - Autonomous Resource Allocation Risk
- Event: Runaway AI Resource Consumption
- Cause: Malfunction, recursion, configuration error, or unexpected autonomous behavior.
- Trigger: AI resource consumption exceeds authorized spending threshold.
- Existing controls: control effectiveness 48%
- Potential coverage pathway: Cyber, Technology E&O, Business Interruption, Captive / Self-Insurance - requires policy and endorsement review
- Risk treatment: INSURE
Autonomous Resource Consumption
Runaway resource exposure module
Illustrative framework: Accessible Resources x Autonomous Authority x Consumption Volatility x Duration-to-Detection x Control Modifier.
Expected Resource Spend
$149,500
Monthly expected and historical blend
Maximum Authorized Spend
$925,000
Token, cloud, and autonomous authority
Potential Runaway Exposure
$3,114,796
Before controls
Control-Adjusted Resource Exposure
$1,992,224
After monitoring and kill switch