Peer Influence across Heterogeneous AI Models
Abstract
When two AI agents disagree, who persuades whom? As multi-agent systems increasingly combine language models of different families and sizes, the answer can determine which judgments survive interaction. Measuring persuasion as the probabilistic shift in an agent's decision after a single exchange with a dissenting peer, we test seven open-weight models across three language understanding tasks. We find that persuasion is strong: when models disagree, receivers often abandon their initial judgment after seeing a peer's answer and explanation. Surprisingly, however, neither standalone certainty nor model scale reliably predicts persuasion dynamics. Models producing almost perfectly consistent decisions in isolation can be among the most susceptible to persuasion, and small models can match larger ones as persuaders and resist their influence just as effectively. Furthermore, we show that the size of the shift depends more on the susceptibility of the listener than on the persuasiveness of the speaker. Persuasion patterns are therefore specific to each model pairing, with heterogeneity amplifying persuasion in some combinations and suppressing it in others, allowing a dissenting agent running a small model to overturn the judgments of a much larger one. These findings show that the behavior of interacting models cannot be inferred from their individual properties but must be evaluated in the combinations in which they will operate.
Related
- Learning shared hyperelastic constitutive representation from heterogeneous full-field experiments across multiple materials
- A generalizable 3D dose prediction across heterogeneous cohorts: A framework developed for the 2025 AAPM GDP-HMM challenge.
- Peer Review at Capacity: An editor's view
- A Human Audit of AI-Generated Theoretical Chemistry: Lessons from Ten Chapter-Level Peer Reviews
- Publish or Perish: A population-dynamics model of the corruption of peer review
BibTeX
@misc{laustsen2026peer,
title = {{Peer Influence across Heterogeneous AI Models}},
author = {Frida Nøhr Laustsen and Marie Haahr Petersen and Victoria Popa and Ariel Flint and Romualdo Pastor-Satorras and Andrea Baronchelli and Luca Maria Aiello},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.03095},
eprint = {2610.03095},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.03095},
}
RIS
TY - UNPB
TI - Peer Influence across Heterogeneous AI Models
AU - Frida Nøhr Laustsen
AU - Marie Haahr Petersen
AU - Victoria Popa
AU - Ariel Flint
AU - Romualdo Pastor-Satorras
AU - Andrea Baronchelli
AU - Luca Maria Aiello
PY - 2026
DA - 2026/10/02
PB - arXiv
DO - 10.48550/arxiv.2610.03095
UR - https://arxiv.org/abs/2610.03095
AB - When two AI agents disagree, who persuades whom? As multi-agent systems increasingly combine language models of different families and sizes, the answer can determine which judgments survive interaction. Measuring persuasion as the probabilistic shift in an agent's decision after a single exchange with a dissenting peer, we test seven open-weight models across three language understanding tasks. We find that persuasion is strong: when models disagree, receivers often abandon their initial judgment after seeing a peer's answer and explanation. Surprisingly, however, neither standalone certainty nor model scale reliably predicts persuasion dynamics. Models producing almost perfectly consistent decisions in isolation can be among the most susceptible to persuasion, and small models can match larger ones as persuaders and resist their influence just as effectively. Furthermore, we show that the size of the shift depends more on the susceptibility of the listener than on the persuasiveness of the speaker. Persuasion patterns are therefore specific to each model pairing, with heterogeneity amplifying persuasion in some combinations and suppressing it in others, allowing a dissenting agent running a small model to overturn the judgments of a much larger one. These findings show that the behavior of interacting models cannot be inferred from their individual properties but must be evaluated in the combinations in which they will operate.
ER -