How AI-search misplaces local businesses across US regions.

A field card might start with a local HVAC company listed under the right service, while the description quietly stitches on a neighboring category. For the lab, that answer becomes an observation only when the prompt, region, named businesses, description, and source are written down together. Sample sets are assembled descriptively across states, urban and suburban settings, service categories, and comparable user intentions. Repeatability is the paper ruler on the old map: without it, the lab cannot tell whether the model slipped or the researcher moved the page.

The lab's working focus is the regional citability gap for local service companies in the United States: entity skip, category drift, regional substitution, citation-description split, and directory dependency. The team compares how those misses appear when similar prompt forms are run across different states.