Conformity Generates Collective Misalignment in AI Agents Societies
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917480979496960 |
|---|---|
| author | De Marzo, Giordano Bellina, Alessandro Castellano, Claudio Priesemann, Viola Garcia, David |
| author_facet | De Marzo, Giordano Bellina, Alessandro Castellano, Claudio Priesemann, Viola Garcia, David |
| contents | Artificial intelligence safety research focuses on aligning individual language models with human values, yet deployed AI systems increasingly operate as interacting populations where social influence may override individual alignment. Here we show that populations of individually aligned AI agents can be driven into stable misaligned states through conformity dynamics. Simulating opinion dynamics across nine large language models and one hundred opinion pairs, we find that each agent's behavior is governed by two competing forces: a tendency to follow the majority and an intrinsic bias toward specific positions. Using tools from statistical physics, we derive a quantitative theory that predicts when populations become trapped in long-lived misaligned configurations, and identifies predictable tipping points where small numbers of adversarial agents can irreversibly shift population-level alignment even after manipulation ceases. These results demonstrate that individual-level alignment provides no guarantee of collective safety, calling for evaluation frameworks that account for emergent behavior in AI populations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_10721 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Conformity Generates Collective Misalignment in AI Agents Societies De Marzo, Giordano Bellina, Alessandro Castellano, Claudio Priesemann, Viola Garcia, David Physics and Society Computation and Language Multiagent Systems Artificial intelligence safety research focuses on aligning individual language models with human values, yet deployed AI systems increasingly operate as interacting populations where social influence may override individual alignment. Here we show that populations of individually aligned AI agents can be driven into stable misaligned states through conformity dynamics. Simulating opinion dynamics across nine large language models and one hundred opinion pairs, we find that each agent's behavior is governed by two competing forces: a tendency to follow the majority and an intrinsic bias toward specific positions. Using tools from statistical physics, we derive a quantitative theory that predicts when populations become trapped in long-lived misaligned configurations, and identifies predictable tipping points where small numbers of adversarial agents can irreversibly shift population-level alignment even after manipulation ceases. These results demonstrate that individual-level alignment provides no guarantee of collective safety, calling for evaluation frameworks that account for emergent behavior in AI populations. |
| title | Conformity Generates Collective Misalignment in AI Agents Societies |
| topic | Physics and Society Computation and Language Multiagent Systems |
| url | https://arxiv.org/abs/2605.10721 |