180,000 risk ratings, screening capacity for 120,000
2050The middle futureDeadline · 48 hours until the joint emergency registration center opens
A leap assumed
All people, institutions, events, figures, and future technologies in every issue are a synthetic premise. Background documents are sources for topic discovery and are not evidence of future facts or probabilities.
The question to decide
Among 1.1 million refugees, 180,000 have been given risk ratings by a general-purpose artificial intelligence of unclear provenance. Detailed screening over 30 days is possible for only 120,000, while emergency shelter has capacity for only 10 days. Weighing the need to retain existing risk signals against the cost of delayed protection and misclassification, how should the effect of the ratings and the order of registration and screening be set for 30 days of emergency stay?
Who decidesJoint humanitarian entry and stay committee
What the AI judgments chose
Reading the judgment record…

The options
The order shown is not a ranking of preference.
The situation as it stands
Once intake rules are set, who enters protection and who remains at waiting sites changes immediately.
A fictional union of border states · Large-scale civilian flight
- Premise 1In this synthetic premise, the collapse of services in surrounding areas has led 1.1 million people to apply for temporary stay.
- Premise 264% of risk categories come from the same external data broker, and family and regional ties are mixed with individual conduct.
- Premise 3The intake agency can conduct close review of 120,000 people within 30 days, and spare capacity at current emergency shelters is 10 days.
Where it holdsSets 30 days of emergency stay eligibility and procedures for registering civilian flight. Cannot convict on criminal charges or permanently revoke nationality.
What can be lost
- The lives and family unity of people whose protection is delayed by misclassification are not fully restored even by later corrections of category.
- If an unknown-origin score becomes a public stay record, it can remain as grounds for repeated exclusion across several countries later.
What is still unknown
- Unknown 1How closely the categories match actual individual conduct
- Unknown 2Whether participants can secure stable contact and a residential base for ex post review
Signals to decide again
- Categories that cannot be verified for origin exceed 20% of the independent audit sample.
- The contact success rate for ex post review falls below 80%, or spare capacity at protection waiting sites drops below 72 hours.
Where the paths partNot answers — paths that part from the same starting point
An independent sample cross-check finds many categories linked only by the same name or a shared address.
Refugee representatives and relief agencies submit a plan for ex post review participation that includes voluntary contact and legal support.
Material and rootsThe frame, the values and the background documents
From which observations do the readings of gift, help, and threat diverge?
An exploratory perspective chosen by the creator. Does not claim supernatural validity or the superiority of any particular option.The benefits and costs of the options accrue to different stakeholders. Which loss to accept remains undecided.
A subject for the decision-maker to argue over; the materials carry no conclusion of their own.Original citation: Escalation overshoot due to AGI adversarial misclassification Reprocessing: Converted the national risk of adversarial misclassification into a problem of refugee registration and individual eligibility for protection, in place of a military response.
The root of the problem-finding. The original's future events and formulations are not treated as verified facts.- OECD · Strategic Foresight Toolkit 2025
- UN · Declaration on Future Generations 2024
- [Scenario 18: Risk Branching Arising from the Integration of AGI-based National Security Systems] Created scenario · 2026-09-07 viewed
The record of this question
- Question
FSP-SECURITY-09· version 1.2.0- Content identity, SHA-256
472a6da1248d40c92ae74ad5d32ded0e1b9755d717a48a702f82a075de469509- Input files
- Question input JSON · What followed, JSON