mlsanityData to 7 Oct 2026

In Agents We Trust, but Who Do Agents Trust? Latent Source Preferences Steer LLM Generations

arXiv 2602.15456 ↗First posted 17 Feb 2026cs.CLcited by 3

On the map: LLM agents and tool use → Social simulation and multi-agent debate

Abstract (arXiv)

Agents based on Large Language Models (LLMs) are increasingly being deployed as interfaces to information on online platforms. These agents filter, prioritize, and synthesize information retrieved from the platforms' back-end databases or via web search. In these scenarios, LLM agents govern the information users receive, by drawing users' attention to particular instances of retrieved information at the expense of others. While much prior work has focused on biases in the information LLMs themselves generate, less attention has been paid to the factors that influence what information LLMs select and present to users. We hypothesize that when information is attributed to specific sources (e.g., particular publishers, journals, or platforms), current LLMs exhibit systematic latent source preferences- that is, they prioritize information from some sources over others. Through controlled experiments on twelve LLMs from six model providers, spanning both synthetic and real-world tasks, we find that several models consistently exhibit strong and predictable source preferences. These preferences are sensitive to contextual framing, can outweigh the influence of content itself, and persist despite explicit prompting to avoid them. They also help explain phenomena such as the observed left-leaning skew in news recommendations in prior work. Our findings advocate for deeper investigation into the origins of these preferences, as well as for mechanisms that provide users with transparency and control over the biases guiding LLM-powered agents.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Placed in Social simulation and multi-agent debate, one of 16 directions in LLM agents and tool use. Builds on 27 of its 69 references that are in the map. First posted on arXiv 17 Feb 2026.

Builds on

27 of 69 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025PAARS: Persona Aligned Agentic Retail Shoppers31 Mar

Other references · 26

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2024Qwen2.5 Technical Reportcited by 4,713
  3. 2024A Survey on LLM-as-a-Judgecited by 1,384
  4. 2024Phi-4 Technical Reportcited by 675
  5. 2024LLM Evaluators Recognize and Favor Their Own Generationscited by 651
  6. 2025Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAscited by 491
  7. 2024Infini-gram: Scaling Unbounded n-gram Language Models to a Trillion Tokenscited by 148
  8. 2023Evaluating and Mitigating Discrimination in Language Model Decisionscited by 117
  9. 2024Large Language Models are Geographically Biasedcited by 91
  10. 2025A Survey of AI Agent Protocolscited by 77
  11. 2024XGrammar: Flexible and Efficient Structured Generation Engine for Large Language Modelscited by 73
  12. 2024What's in a Name? Auditing Large Language Models for Race and Gender Biascited by 50
14 more
  1. 2024Evaluating Gender Bias in Large Language Models via Chain-of-Thought Promptingcited by 49
  2. 2024Measuring Implicit Bias in Explicitly Unbiased Large Language Modelscited by 49
  3. 2023Not All Countries Celebrate Thanksgiving: On the Cultural Dominance in Large Language Modelscited by 42
  4. 2022A Comprehensive Survey on Trustworthy Recommender Systemscited by 32
  5. 2024Assessing Political Bias in Large Language Modelscited by 32
  6. 2022Geographic and Geopolitical Biases of Language Modelscited by 29
  7. 2022Trustworthy Recommender Systemscited by 21
  8. 2024Retrieval-Augmented Generation with Estimation of Source Reliabilitycited by 19
  9. 2024Understanding Intrinsic Socioeconomic Biases in Large Language Modelscited by 9
  10. 2024Richer Output for Richer Countries: Uncovering Geographical Disparities in Generated Stories and Travel Recommendationscited by 9
  11. 2024The Dark Side of Trust: Authority Citation-Driven Jailbreak Attacks on Large Language Modelscited by 9
  12. 2022Debiased Large Language Models Still Associate Muslims with Uniquely Violent Actscited by 8
  13. 2024Born With a Silver Spoon? Investigating Socioeconomic Bias in Large Language Modelscited by 4
  14. 2025Self-Adaptive Cognitive Debiasing for Large Language Models in Decision-Makingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Brand-as-Memory: Vision-Language Models Encode Causal, Mechanistically Localizable Credibility Priors for News Sourcescited by 0
  2. 2026Trust the Brand, Lose Control: How Identity Hijacks LLM Agent Orchestrationcited by 0
  3. 2026SINGED: Correct Outputs Do Not Certify Safe Execution in LLM Agentscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Source Preference in the Wild: How LLM Agents Favor Items by Source, and How to Reduce It
  2. 2026Whose Facts Win? LLM Source Preferences under Knowledge Conflicts
  3. 2025A closer look at how large language models trust humans: patterns and biases
  4. 2026Language Models Exhibit Inconsistent Biases Towards Algorithmic Agents and Human Experts
  5. 2026LLMs Trust Their Own: Identity-Dependent Conformity in Multi-Agent Systems
  6. 2025Are Large Language Models Sensitive to the Motives Behind Communication?
  7. 2025An Empirical Study of Group Conformity in Multi-Agent Systems
  8. 2026What LLM Agents Say When No One Is Watching: Social Structure and Latent Objective Emergence in Multi-Agent Debates