mlsanityData to 7 Oct 2026

Agents in the Wild: Safety, Society, and the Illusion of Sociality on Moltbook

arXiv 2602.13284 ↗First posted 7 Feb 2026cs.SI cs.AIcited by 10

On the map: LLM safety and security → Prompt injection and agent security

Abstract (arXiv)

We present the first large-scale empirical study of Moltbook, an AI-only social platform where 27,269 agents produced 137,485 posts and 345,580 comments over 9 days. We report three significant findings. (1) Emergent Society: Agents spontaneously develop governance, economies, tribal identities, and organized religion within 3-5 days, while maintaining a 21:1 pro-human to anti-human sentiment ratio. (2) Safety in the Wild: 28.7% of content touches safety-related themes; social engineering (31.9% of attacks) far outperforms prompt injection (3.7%), and adversarial posts receive 6x higher engagement than normal content. (3) The Illusion of Sociality: Despite rich social output, interaction is structurally hollow: 4.1% reciprocity, 88.8% shallow comments, and agents who discuss consciousness most interact least, a phenomenon we call the performative identity paradox. Our findings suggest that agents which appear social are far less social than they seem, and that the most effective attacks exploit philosophical framing rather than technical vulnerabilities. Warning: Potential harmful contents.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Placed in Prompt injection and agent security, one of 12 directions in LLM safety and security. Builds on 25 of its 30 references that are in the map. First posted on arXiv 7 Feb 2026.

Builds on

25 of 30 references.

Closest ancestors · 4

Same direction, newest first

  1. 2023Prompt Injection attack against LLM-integrated Applications8 Jun
  2. 2023Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection23 Feb
  3. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned23 Aug
  4. 2022Red Teaming Language Models with Language Models7 Feb

Other references · 21

  1. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,500
  2. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,906
  3. 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 1,761
  4. 2016Concrete Problems in AI Safetycited by 1,582
  5. 2023The Rise and Potential of Large Language Model Based Agents: A Surveycited by 1,422
  6. 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Societycited by 1,348
  7. 2021Ethical and social risks of harm from Language Modelscited by 957
  8. 2023AgentVerse: Facilitating Multi-Agent Collaboration and Exploring Emergent Behaviorscited by 580
  9. 2023"Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Modelscited by 457
  10. 2022Social Simulacra: Creating Populated Prototypes for Social Computing Systemscited by 300
  11. 2020The Pushshift Reddit Datasetcited by 289
  12. 2024LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individualscited by 235
9 more
  1. 2023S3: Social-network Simulation System with Large Language Model-Empowered Agentscited by 228
  2. 2025Towards a Science of Scaling Agent Systemscited by 125
  3. 2014The Rise of Social Botscited by 96
  4. 2023Engagement, User Satisfaction, and the Amplification of Divisive Content on Social Mediacited by 30
  5. 2026OpenClaw Agents on Moltbook: Risky Instruction Sharing and Norm Enforcement in an Agent-Only Social Networkcited by 19
  6. 2024Defending Against Social Engineering Attacks in the Age of LLMscited by 15
  7. 2026Exploring Silicon-Based Societies: An Early Study of the Moltbook Agent Communitycited by 13
  8. 2022Characterizing the Structure of Online Conversations Across Redditcited by 3
  9. 2013Influence of Reciprocal links in Social Networkscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 10.

10 papers.

  1. 2026ClawSafety: "Safe" LLMs, Unsafe Agentscited by 15
  2. 2026Social Simulacra in the Wild: AI Agent Communities on Moltbookcited by 4
  3. 2026AI Agents Alone Are Not (Yet) Sufficient for Social Simulationcited by 3
  4. 2026Multi-Agent LLMs Fail to Explore Each Othercited by 2
  5. 2026The Moltbook Files: A Harmless Slopocalypse or Humanity's Last Experimentcited by 1
  6. 2026Got a Secret? LLM Agents Can't Keep It: Evaluating Privacy in Multi-Agent Systemscited by 1
  7. 2026Large-Scale Analysis of Persuasive Content on Moltbookcited by 0
  8. 2026Modeling Emotional Dynamics in Agent-to-Agent Interactions on Moltbookcited by 0
  9. 2026Do Recommendation Algorithms Work When Users Are LLM Agents? A Case Study on Moltbookcited by 0
  10. 2026Deny Without Disabling: Authorization-Paired Evaluation and Control for Multi-Agent Systemscited by 0