Mutual AI Resilience · MAR-HTP

Compare Italy–Albania support workflows

Compare current practice, structured human coordination, and a proposed central coordinator with the same fictional support demand and resources.

Compare civilian support capacity

Start with a fictional queue of aggregate requests for victim support. See how much work a local support service could complete and how many requests could remain open. Use the same time interval and definition of a completed request throughout.

The three arrangements start with the same caseload and processing capacity. Declared throughput sets the workload result. A request count can include repeat requests.

In the fictional example, every arrangement completes 41 of 42 requests. Open requests at the end of period 4 total 1.

Use fictional totals. The worksheet runs locally in this browser and accepts only aggregate exercise inputs. Actual participation, referral and disclosure require separate review and applicable consent.

Shared requests and staffed capacity

Already waiting at the startwhole requestsExercise length

Use consecutive periods of equal duration, such as days or weeks. Enter the capacity that can be staffed during each period.

Period 1 New support requestswhole requests in this period Common available processing capacitywhole requests in this period

Period 2 New support requestswhole requests in this period Common available processing capacitywhole requests in this period

Period 3 New support requestswhole requests in this period Common available processing capacitywhole requests in this period

Period 4 New support requestswhole requests in this period Common available processing capacitywhole requests in this period

Check corrections in a fictional administrative audit

Use the same fictional batch for each arrangement. Enter the number of summaries needing a correction, counting each summary only once. This can test administrative accuracy without recording any personal information.

Shared audit inputs

Administrative summaries in the batchsummariesSummaries reviewedsummaries; blank if unknownReviewed summaries requiring correctionsummaries; blank if unknown

Enter differing audit observations

Test a change in available capacity

A capacity adjustment describes a different condition. For a fair trial of AI assistance, start with the same staffed limit, task scope, demand, and working time.

Enter differing usable capacities

Review declared process costs and safeguards

Same currency and price basis

Compare costs over the same full workload, defined here as opening requests plus new requests. This is a separate budget exercise.

Setup cost of current arrangementsEUR cents Difference in alternative setup costEUR cents, alternative minus current. Negative means cheaper setup. Current cost per requestEUR cents per request Alternative cost per requestEUR cents per request

Safeguards for a fictional cooperation exercise

These declarations help identify what a demonstration still needs. Checking a box does not confer permission, represent any real person’s consent, or certify a service.

Operator opt-in recorded
The participating support operator has voluntarily opted into the declared cooperation exercise.

Cooperation scope confirmed
A designated reviewer has confirmed the permitted purpose and scope of the fictional exercise.

Human review assigned
A designated human reviews material changes and can stop the exercise.

Independent local fallback tested
The local support process was exercised with the shared service unavailable.

This exercise describes workload and local process constraints. It does not measure a person’s risk, the incidence of trafficking, the outcome of assistance, or the causal effect of AI.

What could change the result?

An early handoff helps only if someone has the staff and capacity to follow up. Compare unresolved requests, review effort, and costs in their own units. Test cooperation under the same deadline and staffing first. Then change an assumption and write down why it changed. A lower backlog in one exercise can result from more capacity and provides no estimate of criminal activity or trafficking prevalence.

Test what happens when coordination stops Estimate separate pilot costs