[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84906-en":3,"doc-seo-84906-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84906,1099514068035,"Ezra","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Large language models create an uneven informational layer over cities","Large language models (LLMs) are emerging as an informational layer over cities, influencing which restaurants people discover, consider, and visit. An audit of restaurant recommendations from three major LLMs across 304 neighborhoods in five U.S. cities, using 320 synthetic user profiles, shows venue fabrication and systematic omission of real places. Fabrication concentrates where digital and physical footprints are weaker and vanishes with verified venue lists, while invisibility persists. The study finds socioeconomic- and user-dependent selectivity and simulates demand shifts that could redirect revenue from chain and quick-service restaurants toward independent, full-service dining, with implications for urban inequality.","arXiv :2607 .06260v 1 [ cs .CY] 7 Jul 2026  \nLarge language models create an uneven informational layer over cities  \nLin Chen 1* , Guangyuan Weng 1,2 , and Esteban Moro 1  \n1 Network Science Institute, Northeastern University, Boston, MA, USA  \n2 Khoury College of Computer Sciences, Northeastern University, Boston, MA, USA  \n* [l.chen2@northeastern.edu](l.chen2@northeastern.edu)  \nABSTRACT  \nLarge language models (LLMs) are emerging as a new informational layer over cities, shaping which places people discover, consider, and ultimately visit. Yet little is known about which places they surface, which they ignore, and whether these patterns vary across communities and users and translate into real-world economic consequences. Here, we audit restaurant recommendations from three major LLMs across 304 neighborhoods in five U.S. cities using 320 synthetic user profiles spanning income, age, sex, and residential status. We find that LLMs both fabricate venues and systematically overlook real ones. Fabrication is concentrated in neighborhoods with weaker digital and physical footprints and disappears when models are provided with verified venue lists. In contrast, invisibility persists: even when choosing from a fixed set of real venues, 47.5% of establishments are never recommended, and 31.9% of these blind spots are shared across all three model families, indicating that uneven visibility reflects not only missing knowledge but also stable patterns of selective attention rooted in shared patterns of visibility rather than model-specific errors. The same selectivity extends to users. Within identical venue pools, higher-income users receive more expensive and less popular venues, while tourists are directed toward costlier but more socially diverse establishments than local residents. Simulating the resulting shifts in consumer demand suggests that widespread reliance on LLM recommendations would redirect visits and revenue away from chain and quick-service restaurants toward independent and full-service dining. Together, our findings show that LLMs act as a selective layer of urban information that unevenly distributes visibility across places and people, with potential consequences for local economies and urban inequality.  \nIntroduction  \nCities are organized not only by their physical infrastructure but also by the information systems that guide how people perceive, navigate, and choose within urban space 1, 2. GPS navigation redistributes traffic across road networks and alters congestion patterns3. Online review platforms reshape consumer visitation and business survival4 , while search engine rankings come to determine the visibility and economic viability of local establishments5. More recently, individuals have begun turning to large language models (LLMs) to choose restaurants, plan trips, and explore unfamiliar neighborhoods6, 7 , often with limited prior knowledge of what is available. From routing between known destinations, to filtering among identified options, to constructing the choice set itself, these information systems have intervened at a progressively earlier stage of urban decision-making. In mediating the earliest of these stages at scale, LLMs increasingly function as a new informational layer superimposed on the physical city. Yet they differ from their predecessors in a structurally consequential way. Search engines and review platforms offer ranked lists whose pagination and scroll depth make the breadth of alternatives visible; a lower-ranked business receives less attention but remains discoverable in principle8. LLM responses do not expose the boundary between what has been included and what has been left out. A venue either appears in the generated response or is entirely absent from the user’s consideration set, with nothing signaling the omission. Understanding the structure of this emerging informational layer, and its potential to reshape which places thrive and which remain unseen, is therefor","cbCaiojbAQBOFZU5","https://ap.wps.com/l/cbCaiojbAQBOFZU5","pdf",22566381,1,27,"English","en",105,"# Abstract\n# Introduction\n## Cities as information systems\n## Unequal urban structure and experiences\n## Biases and open questions","[{\"question\":\"How did the study evaluate what LLMs recommend across cities and neighborhoods?\",\"answer\":\"The researchers audited restaurant recommendations from three major LLMs across 304 neighborhoods in five U.S. cities using 320 synthetic user profiles spanning income, age, sex, and residential status.\"},{\"question\":\"Do the models mainly suffer from missing knowledge, or from a persistent attention pattern?\",\"answer\":\"Fabrication disappears when models receive verified venue lists, suggesting a knowledge-gap component. However, invisibility persists even within fixed real-venue pools, indicating stable selective attention patterns beyond model-specific errors.\"},{\"question\":\"How do recommendations differ for users with the same venue options?\",\"answer\":\"Within identical venue pools, higher-income users are shown more expensive and less popular venues. Tourists are directed to costlier but more socially diverse establishments than local residents.\"}]",1784199282,68,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"large-language-models-create-an-uneven-informational-layer-over-cities","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/large-language-models-create-an-uneven-informational-layer-over-cities/84906/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"How did the study evaluate what LLMs recommend across cities and neighborhoods?","Question",{"text":74,"@type":75},"The researchers audited restaurant recommendations from three major LLMs across 304 neighborhoods in five U.S. cities using 320 synthetic user profiles spanning income, age, sex, and residential status.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"Do the models mainly suffer from missing knowledge, or from a persistent attention pattern?",{"text":79,"@type":75},"Fabrication disappears when models receive verified venue lists, suggesting a knowledge-gap component. However, invisibility persists even within fixed real-venue pools, indicating stable selective attention patterns beyond model-specific errors.",{"name":81,"@type":72,"acceptedAnswer":82},"How do recommendations differ for users with the same venue options?",{"text":83,"@type":75},"Within identical venue pools, higher-income users are shown more expensive and less popular venues. Tourists are directed to costlier but more socially diverse establishments than local residents.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,114,119,122,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":45,"category_name":116,"show_sort_weight":117,"slug":118},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":105,"slug":137},19,"General","general"]