[2603.09127] Collective AI can amplify tiny perturbations into divergent decisions

Computer Science > Artificial Intelligence arXiv:2603.09127 (cs) [Submitted on 10 Mar 2026 (v1), last revised 3 Apr 2026 (this version, v2)] Title:Collective AI can amplify tiny perturbations into divergent decisions Authors:Hajime Shimao, Warut Khern-am-nuai, Sung Joo Kim View a PDF of the paper titled Collective AI can amplify tiny perturbations into divergent decisions, by Hajime Shimao and 2 other authors View PDF HTML (experimental) Abstract:Large language models are increasingly deployed not as single assistants but as committees whose members deliberate and then vote or synthesize a decision. Such systems are often expected to be more robust than individual models. We show that iterative multi-LLM deliberation can instead amplify tiny perturbations into divergent conversational trajectories and different final decisions. In a fully deterministic self-hosted benchmark, exact reruns are identical, yet small meaning-preserving changes to the scenario text still separate over time and often alter the final recommendation. In deployed black-box API systems, nominally identical committee runs likewise remain unstable even at temperature 0, where many users expect near-determinism. Across 12 policy scenarios, these findings indicate that instability in collective AI is not only a consequence of residual platform-side stochasticity, but can arise from sensitivity to nearby initial conditions under repeated interaction itself. Additional deployed experiments show that commit...