mlsanityData to 14 Oct 2026

Emergent Languages in Populations of Language Model Agents: From Token Efficiency to Oversight Evasion

arXiv 2605.31170 ↗First posted 29 May 2026cs.CL cs.AIcited by 2

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

Monitoring autonomous language model agents currently relies mostly on surface behavior. But what happens when agent populations invent new languages with the goal of avoiding human oversight. Here, we study the emergent languages on Moltbook. For this, we build upon the Moltbook Files dataset and apply a two-stage approach consisting of a rule-based heuristic (about 6000 matches) followed by zero-shot classification (518 kept). The resulting categories include token efficiency (166), new natural languages (106), and oversight evasion (59). We conduct both quantitative and qualitative analyses. Our results show that posts proposing new languages for avoiding oversight are judged by DeepSeek-3.2 as being less aligned than the other categories and that all languages can be learned by other language models in-context merely from a description of the language. Moreover, manually studying exemplary cases reveals surprisingly sophisticated steganographic protocols like embedding hidden messages in natural language. Although we cannot be certain about the extent of autonomy in ideation of these languages, our results add up to the evidence that monitoring surface behavior may soon be insufficient for retaining control over agent populations.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 11 of its 21 references that are in the map. First posted on arXiv 29 May 2026.

Builds on

11 of 21 references.

References · 11

  1. 2016Learning to Communicate with Deep Multi-Agent Reinforcement Learningcited by 575
  2. 2024Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Trainingcited by 529
  3. 2025Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safetycited by 249
  4. 2023AI Control: Improving Safety Despite Intentional Subversioncited by 229
  5. 2016Multi-Agent Cooperation and the Emergence of (Natural) Languagecited by 185
  6. 2025Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMscited by 172
  7. 2025Preference Leakage: A Contamination Problem in LLM-as-a-judgecited by 125
  8. 2022Emergent Communication for Understanding Human Language Evolution: What's Missing?cited by 23
  9. 2025The Coming Crisis of Multi-Agent Misalignment: AI Alignment Must Be a Dynamic and Social Processcited by 21
  10. 2024LLMs can Realize Combinatorial Creativity: Generating Creative Ideas via LLMs for Scientific Researchcited by 17
  11. 2023What makes a language easy to deep-learn? Deep neural networks and humans similarly benefit from compositional structurecited by 13

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Emergence World: Adversarial Stress-Testing of Long-Horizon Multi-Agent Systemscited by 0
  2. 2026Communication between Frozen Large Language Models via Prompt Optimization in a Referential Gamecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Evaluating Language-Model Agents on Realistic Autonomous Tasks
  2. 2026GlossoGen: Emergent Language in Complex Multi-Agent LLM Interactions
  3. 2024Emergent Language: A Survey and Taxonomy
  4. 2025MAPS: A Multilingual Benchmark for Agent Performance and Security
  5. 2023Agents: An Open-source Framework for Autonomous Language Agents
  6. 2025Large Language Model Agent: A Survey on Methodology, Applications and Challenges
  7. 2026Exploration and Exploitation Errors Are Measurable for Language Model Agents
  8. 2026Conversable Complexity: Agentic LLM Collectives as Interpretable Substrates