mlsanityData to 7 Oct 2026

LLMs Trust Their Own: Identity-Dependent Conformity in Multi-Agent Systems

arXiv 2609.33495 ↗First posted 27 Sep 2026cs.CL cs.MAcited by 0

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

Abstract (arXiv)

Large language models (LLMs) are increasingly deployed in multi-agent settings, where agents observe and influence one another, making social influence a key dimension of AI behavior and safety. We investigate whether LLMs' responses depend on the social identity of other agents, beyond the effect of their consensus. We construct judgment tasks with a single correct answer, and place models in a multi-agent setting where they receive incorrect answers from other agents whose social identities (AI or human, model family, or an arbitrary minimal group) are either shared with or distinct from their own. Across 12 open-weights models and nine tasks, we find a bidirectional effect of group identity on conformity to incorrect answers: in-group consensus increases conformity (in-group favoritism), whereas out-group consensus decreases it (out-group divergence). Unlike humans, for whom one ally breaking the consensus sharply reduces conformity, models are unmoved by an ally from the majority's group. Worse, a correct ally from the opposing group intensifies this bidirectional effect. Chain-of-Thought reasoning suppresses most of these effects, yet an in-group ally still reduces conformity to an incorrect out-group majority. Labeling peers as safety-aligned shifts overall conformity but leaves in-group favoritism and out-group divergence intact. These results show that group identity shapes how LLMs aggregate information across agents, independently of its correctness, and identify a manipulation surface for multi-agent AI systems.

Key facts. No citing papers in the map yet (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 31 of its 58 references that are in the map. First posted on arXiv 27 Sep 2026.

Builds on

31 of 58 references.

Closest ancestors · 10

Same direction, newest first

  1. 2026Toward a social psychology of AI: language-model agents reproduce human-like minimal-group bias29 Jul
  2. 2026Language model agents show in-group trust bias invisible to standard behavioural audits27 May
  3. 2026Truth or Tribe: How In-group Favoritism Prioritize Facts in Persona Agents2 May
  4. 2026Large Language Models Exhibit Normative Conformity21 Apr
  5. 2026Conformity and Social Impact on AI Agents8 Jan
  6. 2026When Agents See Humans as the Outgroup: Belief-Dependent Bias in LLM-Powered Agents1 Jan
  7. 2025From Single to Societal: Analyzing Persona-Induced Bias in Multi-Agent Interactions14 Nov
  8. 2025LLMs Can't Handle Peer Pressure: Crumbling under Multi-Agent Social Interactions24 Aug
  9. 2024Exploring Prosocial Irrationality for LLM Agents: A Social Cognition View23 May
  10. 2024Persona Inconstancy in Multi-Agent LLM Collaboration: Conformity, Confabulation, and Impersonation6 May

Other references · 21

  1. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  2. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,710
  3. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,767
  4. 2024MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmarkcited by 1,758
  5. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,081
  6. 2023Towards Understanding Sycophancy in Language Modelscited by 988
  7. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 805
  8. 2023Whose Opinions Do Language Models Reflect?cited by 759
  9. 2022Out of One, Many: Using Language Models to Simulate Human Samplescited by 737
  10. 2024LLM Evaluators Recognize and Favor Their Own Generationscited by 651
  11. 2025Reasoning Models Don't Always Say What They Thinkcited by 388
  12. 2023Generative Language Models Exhibit Social Identity Biasescited by 107
9 more
  1. 2025Do as We Do, Not as You Think: the Conformity of Large Language Modelscited by 48
  2. 2024Conformity in Large Language Modelscited by 30
  3. 2024AI-AI Bias: large language models favor communications generated by large language modelscited by 25
  4. 2025Sycophancy Mitigation Through Reinforcement Learning with Uncertainty-Aware Adaptive Reasoning Trajectoriescited by 10
  5. 2026Easier to Mislead Than to Correct: Harmful and Beneficial Revision in LLM Conformitycited by 6
  6. 2024I Am Not Them: Fluid Identities and Persistent Out-group Bias in Large Language Modelscited by 5
  7. 2025Disentangling the Drivers of LLM Social Conformity: An Uncertainty-Moderated Dual-Process Mechanismcited by 5
  8. 2026From Simulation to Enaction: Post-trained language models recognize and react to their own generationscited by 3
  9. 2026Aligned Alone, Misaligned Together: Forecasting Adversarial Capture in LLM Agent Populationscited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Conformity Dynamics in LLM Multi-Agent Systems: The Roles of Topology and Self-Social Weighting
  2. 2025An Empirical Study of Group Conformity in Multi-Agent Systems
  3. 2025The Social Laboratory: A Psychometric Framework for Multi-Agent LLM Evaluation
  4. 2026Prompted Identity Degrades Cooperation in Multi-Agent LLM Systems
  5. 2026"Who Am I, and Who Else Is Here?" Behavioral Differentiation Without Role Assignment in Multi-Agent LLM Systems
  6. 2026Not All Flips Are Conformity: Decomposing Stance Convergence in Multi-Agent LLM Debate
  7. 2026Collective Opinion Dynamics in Structured LLM Populations
  8. 2026Social Networks of LLM Agents