Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off
Public institutions increasingly use large language models (LLMs) to answer citizens' questions, often pairing a curated knowledge base with live web search, yet whether the sources behind these answers can be trusted has received little empirical scrutiny. We report a pre-launch expert evaluation of Evrópuvefur, an independent, government-funded service run by the University of Iceland that answers questions about the European Union, conducted as Iceland prepared for its referendum of 29 August 2026 on whether to resume EU accession talks. Five domain experts produced 551 evaluations of 449 A
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 58%Supporting Europe’s work in ensuring a trustworthy AI ecosystem →
- PossiblePossibly related (embedding) · 54%Why Agencies That Ignore AI-Powered Search Tools Are Already Falling Behind Their Competitors - Resident Magazine →
- PossiblePossibly related (embedding) · 54%A new era for AI Search →
- PossiblePossibly related (embedding) · 53%arc53/DocsGPT →
- PossiblePossibly related (embedding) · 53%mlimarenko/IronRAG →
- LinkedLinked via arxiv author · 85%Hafsteinn Einarsson →
“Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off”
- LinkedLinked via arxiv author · 85%Hafsteinn Birgir Einarsson →
“Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off”
- LinkedLinked via arxiv author · 85%Jón Gunnar Ólafsson →
“Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off”