ORS · Resilience dimension
AI and Agentic Resilience
AI does not deserve its own scorecard beside the institution's resilience position. It belongs inside it. This dimension aggregates fifteen inputs and reports one thing: the effect AI and autonomous systems have had on the institution's ability to keep operating.
Fifteen inputs
Input names are published. Weights, thresholds and the grading rubric are institutional methodology and are not.
01Model dependency
02Agent autonomy
03Tool access
04Data permissions
05Human oversight
06Model drift
07Agent drift
08Provider concentration
09Third-party dependency
10Cyber exposure
11Incident history
12Recovery capability
13Fallback model
14Evidence freshness
15Control effectiveness
What the board receives
Not this
AI Governance Score: 82
A number with no consequence attached. It does not tell the board what changed, or what to do about it.
This
AI has moved the institution's resilience position, and here is the dependency that did it
Traceable to a named control, a named provider and a named owner, with the evidence behind each.