A privacy brief exposing a second census built from travel, housing, clinic, and observatory records without the knowledge of the people being counted.
The Veyr Census Office commissioned House Nhal-Serat to estimate how many people lived in the South Basin without current identity records. The public purpose was reasonable. Emergency food, clinic supplies, water, and evacuation plans all depend on knowing how many people may need them.
The office did not conduct a public count. It built an inferred one.
Nhal-Serat analysts combined ferry entries, rent payments, bathhouse attendance, clinic visits, school meals, work shifts, market purchases, mirror-street traffic counts, and the times lights appeared in residential windows. The model assigned each observed person a probable home, household, employment, health risk, debt status, and chance of leaving the district.
The results were delivered as a private supplement to the official census. Residents could not see it or correct it. Most did not know it existed.
The supplement first became public through the case of Talin Marr.
Talin lived alone and regularly cared for a friend recovering from surgery. The friend's building appeared in travel records more often than Talin's registered home. The model reassigned Talin to the friend's household and marked the arrangement as a probable unregistered partnership.
That inference moved Talin's food allocation, linked the two people's debt risks, and caused a work-travel renewal to be delayed until the household discrepancy was reviewed. The friend learned of the inferred partnership from the refusal notice. Neither person had been asked.
Other errors followed. Night workers were placed in buildings where they ate before shift. Children who moved between two parents were assigned to whichever doorway the street mirror observed most clearly. Clinic visitors were grouped by shared symptoms and treated as households. People using sanctuary routes were placed back into the homes they had fled because old rent records outweighed protected travel.
The Veyr office argued that the supplement was only a planning tool. House Nhal-Serat argued that every entry carried a probability and therefore made no claim of fact.
The brief found that both defenses failed in practice.
The supplement affected food, permits, debt review, travel, and investigation. An uncertain record used to make a decision is not harmless because its uncertainty is visible only to the analyst. A private planning tool becomes a second census when public offices act upon it.
The project also changed the meaning of consent. A person who entered a ferry expected their route to be recorded. A clinic patient expected a treatment record. A tenant expected a rent account. None had agreed for those separate records to be combined into a forecast of family, debt, health, and future movement.
The brief ordered every decision based on the supplement to identify it as a source. Residents received access to the entries attached to their names, the observations used, the confidence assigned, and the offices that received the result. They could correct facts, challenge inferences, and demand removal of protected sanctuary data.
The planning office could retain anonymous population totals where emergency supply required them. It could not preserve a hidden personal record merely because the names were useful for other work.
Talin's official household was restored. The friend's medical information was removed from the census supplement. The work-travel renewal was granted. No ruling was made about the nature of their relationship because no public office needed to decide it.
The South Basin still needs an accurate count. Accuracy does not require building a private life behind every name. Ask people. Publish the purpose. Protect the answers. Do not place a second census behind a mirror and call it weather.
