# Desk log AIPA-HOLD-20260823

**Status:** First-party illustrative sanitized composite; not employee, customer, research, benchmark, or independently validated data  
**Run date:** 2026-08-23  
**Evidence checked:** 2026-08-31  
**Operator:** InfiniSynapse Data Team  
**Attestor:** William Zhu, InfiniSynapse cofounder ([GitHub @allwefantasy](https://github.com/allwefantasy))

## Experience and accuracy record

The operator tested an aggregate EMEA tenure-band request with an n-under-five rule and complementary suppression. The retained table showed 6.1% and 8.4% for the prior and current quarter in the 13–24-month band. Two small cells and one parent complement were withheld.

A “score the roster” request and predicted-quit narrative were rejected. No names, identifiers, comments, person-level scores, source counts, or customer outcomes are represented.

The [published CSV](https://infinisynapse.com/blog-media/ai-people-analytics/downloads/band-output-AIPA-HOLD-20260823.csv) exposes two rates and three blank hold records. The [verification script](https://infinisynapse.com/blog-media/ai-people-analytics/downloads/verify-AIPA-HOLD-20260823.py) checks only those published outputs; unavailable denominators prevent independent rate recomputation.

## External context

- [NIST Privacy Framework](https://www.nist.gov/privacy-framework) (retrieved 2026-08-31)
- [U.S. EEOC AI initiative](https://www.eeoc.gov/ai) (retrieved 2026-08-31)
- [UK ICO employment data guidance](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/employment/) (retrieved 2026-08-31)

These authorities did not review the run or product.