mlsanityData to 14 Oct 2026

Position: AI Is Not Ready for Strategic Conflicts

arXiv 2609.16189 ↗First posted 31 Jul 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Open-ended strategic wargames are high-stakes LM-based social simulations: they model adversaries, institutions, escalation, plan brittleness, doctrine, and crisis response. Language models (LMs) are attractive because they can play agents, generate scenario branches, adjudicate ambiguous actions, and summarize lessons, but the same affordances make open-ended roles dangerous: model language determines both what an actor attempts and what becomes simulated reality. This position paper argues that no LM-enabled wargame should inform planning, doctrine, policy, or crisis response without an auditable safety case, and that the proper use of open-ended wargames today is to stress-test decision-influencing LM agents. We identify five failure modes: decision laundering, adjudication opacity, role collapse, escalation-through-adjudication, and failure of strategic imagination. Ordinary benchmarks cannot establish safety for these settings. Wargames can expose failures as stress tests; they are not themselves safety cases for consequential use.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 34 of its 85 references that are in the map. First posted on arXiv 31 Jul 2026.

Builds on

34 of 85 references.

References · 34

  1. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
  2. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,420
  3. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,122
  4. 2021A General Language Assistant as a Laboratory for Alignmentcited by 1,055
  5. 2023Towards Understanding Sycophancy in Language Modelscited by 1,040
  6. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 703
  7. 2024LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methodscited by 465
  8. 2025Persona Vectors: Monitoring and Controlling Character Traits in Language Modelscited by 345
  9. 2019Measurement and Fairnesscited by 308
  10. 2024AI Agents That Mattercited by 219
  11. 2024BALROG: Benchmarking Agentic LLM and VLM Reasoning On Gamescited by 146
  12. 2024Analyzing the Generalization and Reliability of Steering Vectorscited by 143
22 more
  1. 2022Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learningcited by 131
  2. 2025NoLiMa: Long-Context Evaluation Beyond Literal Matchingcited by 126
  3. 2024ForecastBench: A Dynamic Benchmark of AI Forecasting Capabilitiescited by 110
  4. 2024Systematic Biases in LLM Simulations of Debatescited by 99
  5. 2024GameBench: Evaluating Strategic Reasoning Abilities of LLM Agentscited by 95
  6. 2020Human-centered Explainable AI: Towards a Reflective Sociotechnical Approachcited by 90
  7. 2024Escalation Risks from Language Models in Military and Diplomatic Decision-Makingcited by 63
  8. 2024Safety Cases: How to Justify the Safety of Advanced AI Systemscited by 48
  9. 2022Dungeons and Dragons as a Dialog Challenge for Artificial Intelligencecited by 35
  10. 2025SPIN-Bench: How Well Do LLMs Plan Strategically and Reason Socially?cited by 24
  11. 2023CALYPSO: LLMs as Dungeon Masters' Assistantscited by 23
  12. 2024Human vs. Machine: Behavioral Differences Between Expert Humans and Language Models in Wargame Simulationscited by 21
  13. 2024Measuring Free-Form Decision-Making Inconsistency of Language Models in Military Crisis Simulationscited by 11
  14. 2025Evaluating LLM Agent Collusion in Double Auctionscited by 11
  15. 2025Too Good to be Bad: On the Failure of LLMs to Role-Play Villainscited by 11
  16. 2024Why Would You Suggest That? Human Trust in Language Model Responsescited by 9
  17. 2024Open-Ended Wargames with Large Language Modelscited by 8
  18. 2021Detecting and Adapting to Novelty in Gamescited by 7
  19. 2025ZeroSumEval: An Extensible Framework For Scaling LLM Evaluation with Inter-Model Competitioncited by 5
  20. 2023Self Generated Wargame AI: Double Layer Agent Task Planning Based on Large Language Modelcited by 2
  21. 2024A Mixture-of-Experts Approach to Few-Shot Task Transfer in Open-Ended Text Worldscited by 2
  22. 2025Autonomous generation of different courses of action in mechanized combat operationscited by 1

Built on by

Not cited yet.