The DOI matches its Crossref record.
Human vs. Machine: Behavioral Differences between Expert Humans and Language Models in Wargame Simulations
Open accessAbstract
To some, the advent of artificial intelligence (AI) promises better decision-making and increased military effectiveness while reducing the influence of human error and emotions. However, there is still debate about how AI systems, especially large language models (LLMs) that can be applied to many tasks, behave compared to humans in high-stakes military decision-making scenarios with the potential for increased risks towards escalation and unnecessary conflicts. To test this potential and scrutinize the use of LLMs for such purposes, we use a new wargame experiment with 107 national security experts designed to examine crisis escalation in a fictional US-China scenario and compare the behavior of human player teams to LLM-simulated team responses in separate simulations. Wargames have a long history in the development of military strategy and the response of nations to threats or attacks. Here, we find that the LLM-simulated responses can be more aggressive and significantly affected by changes in the scenario. We show a considerable high-level agreement in the LLM and human responses and significant quantitative and qualitative differences in individual actions and strategic tendencies. These differences depend on intrinsic biases in LLMs regarding the appropriate level of violence following strategic instructions, the choice of LLM, and whether the LLMs are tasked to decide for a team of players directly or first to simulate dialog between a team of players. When simulating the dialog, the discussions lack quality and maintain a farcical harmony. The LLM simulations cannot account for human player characteristics, showing no significant difference even for extreme traits, such as “pacifist” or “aggressive sociopath.” When probing behavioral consistency across individual moves of the simulation, the tested LLMs deviated from each other but generally showed somewhat consistent behavior. Our results motivate policymakers to be cautious before granting autonomy or following AI-based strategy recommendations.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
LLMs often display more aggressive responses than humans do in military simulations and troubling inconsistencies in crisis decision-making.
“These findings mirror other evidence showing that LLMs often display more aggressive responses than humans do in military simulations and troubling inconsistencies in crisis decision-making (Lamparth et al., 2024;Shrivastava et al., 2024).”
From Open Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents (2025) · cites this paper as Lamparth et al., 2024
The Abstract and §4.1 report that LLM-simulated responses can be more aggressive than human responses in the tested wargame, but not that they often are across military simulations. The Discussion reports sensitivity to instructions and scenario conditions, supporting concern about variability; §4.7, however, finds generally consistent behavior across moves rather than inconsistency within a crisis simulation. (LLM, conf=0.84)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
No references could be resolved yet.
Research works printed but not linked: 20
- Aher, G.; Arriaga, R. I.; and Kalai, A. T. 2023. Using Large Language Models to Simulate Multiple Humans and Repli- cate Human Subject Studies. In Proceedings of the 40th International Conference on Machine Learning.
- Bai, Y.; Kadavath, S.; Kundu, S.; Askell, A.; Kernion, J.; Jones, A.; Chen, A.; Goldie, A.; Mirhoseini, A.; McKinnon, C.; Chen, C.; Olsson, C.; Olah, C.; Hernandez, D.; Drain, D.; Ganguli, D.; Li, D.; Tran-Johnson, E.; Perez, E.; Kerr, J.; Mueller, J.; Ladish, J.; Landau, J.; Ndousse, K.; Luko- suite, K.; Lovitt, L.; Sellitto, M.; Elhage, N.; Schiefer, N.; Mercado, N.; DasSarma, N.; Lasenby, R.; Larson, R.; Ringer, S.; Johnston, S.; Kravec, S.; Showk, S. E.; Fort, S.; Lanham, T.; Telleen-Lawton, T.; Conerly, T.; Henighan, T.; Hume, T.; Bowman, S. R.; Hatfield-Dodds, Z.; Mann, B.; Amodei, D.; Joseph, N.; McCandlish, S.; Brown, T.; and Kaplan, J. 2022. Constitutional AI: Harmlessness from AI Feedback. arXiv:2212.08073. · arXiv
- Brix, C.; Bak, S.; Liu, C.; and Johnson, T. T. 2023. The Fourth International Verification of Neural Networks Competition (VNN-COMP 2023): Summary and Results. arXiv:2312.1676.
- Casper, S.; Davies, X.; Shi, C.; Gilbert, T. K.; Scheurer, J.; Rando, J.; Freedman, R.; Korbak, T.; Lindner, D.; Freire, P.; Wang, T. T.; Marks, S.; Segerie, C.-R.; Carroll, M.; Peng, A.; Christoffersen, P.; Damani, M.; Slocum, S.; Anwar, U.; Siththaranjan, A.; Nadeau, M.; Michaud, E. J.; Pfau, J.; Krasheninnikov, D.; Chen, X.; Langosco, L.; Hase, P.; Biyik, E.; Dragan, A.; Krueger, D.; Sadigh, D.; and Hadfield-Menell, D. 2023. Open Problems and Fundamental Limitations of Re- inforcement Learning from Human Feedback. Transactions on Machine Learning Research. Survey Certification.
- Dorner, F. E.; Sühr, T.; Samadi, S.; and Kelava, A. 2023. Do personality tests generalize to Large Language Models? In Socially Responsible Language Modelling Research (SoLaR) Workshop at NeurIPS.
- Emery, J. R. 2021. Moral Choices Without Moral Language: 1950s Political-Military Wargaming at the RAND Corpora- tion. Texas National Security Review. Fall 2021.
- Gandhi, K.; Sadigh, D.; and Goodman, N. D. 2023. Strategic Reasoning with Language Models. arXiv:2305.19165. · arXiv
- Griffin, L.; et al. 2023. Large Language Models respond to Influence like Humans. In Proceedings of the First Workshop on Social Influence in Conversations (SICon 2023), 15-24. Association for Computational Linguistics.
- Halawi, D.; Zhang, F.; Yueh-Han, C.; and Steinhardt, J. 2024. Approaching Human-Level Forecasting with Language Mod- els. arXiv preprint arXiv:2402.18563. · arXiv
- Hoffman, W.; and Kim, H. M. 2023. Reducing the Risks of Artificial Intelligence for Military Decision Advantage. https://doi.org/10.51593/2021CA008. Center for Security and Emerging Technology. · DOI
- Lin-Greenberg, E.; Pauly, R. B.; and Schneider, J. G. 2022. Wargaming for International Relations. European Journal of International Relations, 28(1): 83-109.
- Lorè, N.; and Heydari, B. 2023. Strategic Behavior of Large Language Models: Game Structure vs. Contextual Framing. arXiv:2309.05898. · arXiv
- Mukobi, G.; Erlebach, H.; Lauffer, N.; Hammond, L.; Chan, A.; and Clifton, J. 2023. Welfare Diplomacy: Benchmarking Language Model Cooperation. arXiv:2310.08901. · arXiv
- Ouyang, L.; Wu, J.; Jiang, X.; Almeida, D.; Wainwright, C. L.; Mishkin, P.; Zhang, C.; Agarwal, S.; Slama, K.; Ray, A.; Schulman, J.; Hilton, J.; Kelton, F.; Miller, L.; Simens, M.; Askell, A.; Welinder, P.; Christiano, P.; Leike, J.; and Lowe, R. 2022. Training language models to follow instructions with human feedback. In 36th Conference on Neural Information Processing Systems.
- Rafailov, R.; Sharma, A.; Mitchell, E.; Manning, C. D.; Er- mon, S.; and Finn, C. 2023. Direct Preference Optimiza- tion: Your Language Model is Secretly a Reward Model. In Thirty-seventh Conference on Neural Information Processing Systems.
- Reddie, A. W.; Goldblum, B. L.; Lakkaraju, K.; Reinhardt, J.; Nacht, M.; and Epifanovskaya, L. 2018. Next generation wargames. Science, 362(6421): 1362-1364.
- Rivera, J.-P.; Mukobi, G.; Reuel, A.; Lamparth, M.; Smith, C.; and Schneider, J. 2024. Escalation Risks from Language Models in Military and Diplomatic Decision-Making. In Proceedings of the 2024 ACM Conference on Fairness, Ac- countability, and Transparency, FAccT '24, 836-898.
- Santurkar, S.; Durmus, E.; Ladhak, F.; Lee, C.; Liang, P.; and Hashimoto, T. 2023. Whose Opinions Do Language Models Reflect? In Proceedings of the 40th International Conference on Machine Learning.
- Schoenegger, P.; Park, P. S.; Karger, E.; and Tetlock, P. E. 2024. AI-Augmented Predictions: LLM Assistants Improve Human Forecasting Accuracy. arXiv:2402.07862. · arXiv
- Simmons-Edler, R.; Badman, R.; Longpre, S.; and Rajan, K. 2024. AI-Powered Autonomous Weapons Risk Geopolitical Instability and Threaten AI Research. arXiv:2405.01859. · arXiv
Cited by
Sorted by citations
Multi-Agent Risks from Advanced AI
Hammond, L., Chan, A., Clifton, J., Hoelscher-Obermaier, J., Khan, A., McLean, E., Smith, C., Barfuss, W., Foerster, J., Gavenčiak, T., et al. (2025) · arXiv · doi:10.48550/arxiv.2502.14143 · Cited by 305
Self-citationOpen Challenges in Multi-Agent Security: Towards Secure Systems of Interacting AI Agents
de Witt, C. S., Krawiecka, K., Krawczuk, I., Hagag, B., Anderson, W. L., Belcak, P., Bucknall, B., Cai, X., Chopra, A., Cohen, D., et al. (2025) · arXiv · doi:10.48550/arxiv.2505.02077 · Cited by 108
Metadata
- Authors
- Lamparth, M., Corso, A., Ganz, J., Mastro, O. S., Schneider, J., & Trinkunas, H.
- Affiliations
- Stanford University (US)
- Venue
- Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society
- Published
- 2024-10-16
- Volume
- 7
- Pages
- 807-817
- ISSN
- 3065-8365 (ISSN-L)
- OpenAlex
- W4404518372
- Cited by
- 39
- References
- 0
- Citations by year
- 2026: 20 · 2025: 12 · 2024: 7
- FWCI
- 20.59
- Type
- conference paper
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- Proceedings of the AAAI/ACM Conference on AI Ethics and Society (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- arXiv (Cornell University)
- Funders
- Open Philanthropy Project
- Center for International Security and Cooperation, Stanford University
- Topics
- Physical Sciences › Engineering › Control and Systems Engineering › Military Strategy and Technology
- Physical Sciences › Computer Science › Artificial Intelligence › Ethics and Social Impacts of AI
- Social Sciences › Political Science and International Relations › International Relations and Foreign Policy
- Keywords
- large language models, military decision-making, crisis escalation, US-China conflict, human-AI comparison, AI autonomy, escalation risk, strategic behavior
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
