{
  "id": 6207,
  "url": "https://arxiv.org/abs/2604.06820v1",
  "title": "Beyond Surface Judgments: Human-Grounded Risk Evaluation of LLM-Generated Disinformation",
  "summary": "Large language models (LLMs) can generate persuasive narratives at scale, raising concerns about their potential use in disinformation campaigns. Assessing this risk ultimately requires understanding how readers receive such content. In practice, however, LLM judges are increasingly used as a low-cost substitute for direct human evaluation, even though whether they faithfully track reader responses remains unclear. We recast evaluation in this setting as a proxy-validity problem and audit LLM ju",
  "authors": "Zonghuan Xu, Xiang Zheng, Yutao Wu, Xingjun Ma",
  "category": "research",
  "topics": "misinformation,transparency",
  "orgs": null,
  "regions": null,
  "published_at": "2026-04-08T08:37:37.000Z",
  "fetched_at": "2026-07-14T16:32:20.054Z",
  "source_slug": "arxiv-ethics",
  "source_name": "arXiv",
  "source_homepage": "https://arxiv.org",
  "ethics_ai_record_url": "https://ethics.ai/record/6207",
  "original_url": "https://arxiv.org/abs/2604.06820v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}