Curated retrieval versus open web search in public AI information services: a coverage-trust trade-off
arXiv:2607.05217v2 Announce Type: replace-cross Abstract: 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 Evr\'opuvefur, 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
The increasing deployment of LLMs by public institutions mandates immediate scrutiny of their information veracity, especially with critical geopolitical events like the Icelandic EU referendum approaching.
This paper highlights the critical trade-off between coverage and trust in public AI information services, directly impacting citizen engagement, policy formation, and democratic processes.
The empirical evaluation provides a framework for understanding and mitigating risks associated with AI-powered public information, moving beyond theoretical concerns to practical implementation challenges.
- · Public institutions with strong data governance
- · Citizens seeking reliable information
- · AI ethicists and trust researchers
- · Public institutions deploying AI without careful validation
- · AI systems prioritizing breadth over accuracy
- · Misinformation campaigns
Public institutions will likely adopt more rigorous testing and validation protocols for AI-driven information services.
Increased demand for curated, trusted knowledge bases and a potential decrease in reliance on broad open web search for sensitive public inquiries.
The definition of 'public trust' in AI will evolve, potentially leading to new regulatory frameworks and industry standards for AI transparency and accountability in public services.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL