12,445 papers · continuously updated · last export: 10 Aug 2026livingmeta.ai
← Browse all papers
Research theme

Epistemic Risk

The Epistemic Risk theme comprises 539 papers in this corpus published between 2004 and 2026. Work here is dominated by Conceptual, Empirical Study, Experimental. 5 open research gaps have been surfaced in this area.

Methodology profile

  • Conceptual175 (32%)
  • Empirical Study70 (13%)
  • Experimental58 (11%)
  • Literature Review50 (9%)
  • Position Paper48 (9%)
  • Content Analysis38 (7%)

Research domains

  • Bias & Epistemic Risk283 (53%)
  • Research Integrity197 (37%)
  • AI Governance19 (4%)
  • Human-AI Collaboration17 (3%)
  • Knowledge Synthesis8 (1%)
  • Data Analysis5 (1%)

Frequent sub-topics

LLM personality, language ideology propagation, and societal bias · 1AI hallucination and epistemic authority in information systems · 1judgment-embedded AI outputs and professional decision-making · 1AI-generated image detection and authenticity judgment · 1Indoctrination risk in using large language models for research · 1AI-generated deception detection and cognitive offloading · 1toxic algospeak and content moderation with AI · 1credibility assessment in synthetic media · 1

Open research gaps

  • The authors identify a gap in understanding the qualities of Big Data beyond quantitative aspects, noting that "we lack understanding about what are the qualities of Big Data that may contribute to th
  • The study identifies the need for research on designing fairer multilingual academic evaluation systems and for better understanding the limitations of GenAI as scholarly evaluation infrastructure.
  • The paper identifies the gap between human epistemic capabilities in scientific dispute and current AI system limitations, noting the need for enhanced frameworks to support adversarial reasoning whil
  • The paper identifies the need for critical examination of Big Data's implications across multiple domains: privacy protection, marketing practices, political surveillance, research methodology, and th
  • The authors identify the need for understanding reliability boundaries of self-verification in medical AI systems and highlight gaps in deploying these mechanisms in real clinical settings versus clea

Representative papers

← Browse all papersPriority research agenda →