Abstract

LLM-based social simulations can make a generated transcript look like a single behavioral signal, but the model behind that transcript may be doing several different jobs: choosing what an actor says or does, deciding what happens after an action, or both. The difference matters especially in open-ended wargames, where models are prized for handling unusual actions and ambiguous consequences. We report a scoping review of 223 de-duplicated AI-in-wargames and strategic-simulation papers retrieved through May 1, 2026, describing each simulation by its model-control profile: whether the language model has open-ended control over player actions, adjudication, or both. Only 20 of 223 studies (≈9%) give language models both roles. Before treating LM outputs as social simulations, researchers need to know how much creative control the model has over actions and consequences. For open-ended simulations, fidelity depends not only on whether agents behave plausibly, but also on whether language models can reliably act as adjudicators or world models.

Where the Freedom Sits

The rubric separates what the protocol permits a player to propose from how it permits an adjudicator to determine consequences. These are independent choices, even when both roles use language.

The 223 studies were retrieved through May 1, 2026. Counts describe the protocols in the reviewed literature, not model performance or the quality of a particular playthrough. Only 20 studies use protocols that permit both open-ended player actions and open-ended adjudication.

Cite This Paper

@inproceedings{matlin2026wargames,
  title     = {Shall We Play a Game? Language Models for Open-ended Wargames},
  author    = {Matlin, Glenn and Song, Isaac and Hao, Yixiong and Mahajan, Parv and Montoya, Evan and Bard, Ryan and Topp, Stuart R. and Zang, Anthony Wen-Ming and Parwani, Mohammed Rehan and Shetty, Soham and Riedl, Mark},
  year      = {2026},
  booktitle = {Social Simulation with LLMs: Fidelity in Applications Workshop at the 3rd Conference on Language Models (COLM)},
  note      = {arXiv:2509.17192v3; earlier version at the Wordplay: When Language Meets Games Workshop, EMNLP 2025}
}