Safety
Position: AI Is Not Ready for Strategic Conflicts
arXiv:2609.16189v1 Announce Type: new Abstract: Open-ended strategic wargames are high-stakes LM-based social simulations: they model adversaries, institutions, escalation, plan brittleness, doctrine,
arXiv:2609.16189v1 Announce Type: new Abstract: 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.
Related
- Position: Safety and Fairness in Agentic AI Depend on Interaction Topology, Not on Model Scale or Alignment
- Benchmarking Misuse Mitigation Against Covert Adversaries
- Deterministic Decisions for High-Stakes AI. A Zero-Egress Pipeline with the Deployability of RAG and the Accuracy of Machine Learning
- One Step to the Side: Why Defenses Against Malicious Finetuning Fail Under Adaptive Adversaries
- OpenEvoShield: Dual Non-Stationary Continual Defense for Open-World Multi-Agent System Attacks
Source: arXiv cs.AI | 2026-09-16