Scaffolding Large Language Models (LLMs) into multi-agent systems often improves performance on complex tasks, but the safety impact of such scaffolds has not been thoroughly explored. We introduce AgentBreeder, a framework for multi-objective self-improving evolutionary search over scaffolds. We evaluate discovered scaffolds on widely recognized reasoning, mathematics, and safety benchmarks and compare them with popular baselines. In "blue" mode, we see a 79.4% average uplift in safety benchmark performance while maintaining or improving capability scores. In "red" mode, we find adversarially weak scaffolds emerging concurrently with capability optimization. Our work demonstrates the risks of multi-agent scaffolding and provides a framework for mitigating them. Code is available at https://github.com/jrosseruk/AgentBreeder.
@article{arxiv.2502.00757,
title = {AgentBreeder: Mitigating the AI Safety Risks of Multi-Agent Scaffolds via Self-Improvement},
author = {J Rosser and Jakob Foerster},
journal= {arXiv preprint arXiv:2502.00757},
year = {2025}
}