Internal vs. External: Comparing Deliberation and Evolution for Multi-Agent Constitutional Design
Multi-agent AI systems need behavioral constitutions, but it is unresolved whether such rules should emerge internally through agent self-governance or be discovered externally through optimization. We present the first controlled comparison of internal deliberation and external evolution across three social environments: a coordination grid-world, an iterated public goods game, and a bilateral trading market. Across 180 simulation runs, evolution significantly outperforms deliberation in collec
Record details
Published: 9 May 2026
Source: arXiv
Category: Research
Topics: Regulation · Agents & autonomy · Environment
Retrieved: 14 July 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
A Multi-Level Agent-Based Architecture for Climate Governance Integrating Cognitive and Institutional Dynamics
arXiv · 8 May 2026
Prediction and Empowerment: A Theory of Agency through Bridge Interfaces
arXiv · 7 May 2026
LLM-X: A Scalable Negotiation-Oriented Exchange for Communication Among Personal LLM Agents
arXiv · 12 May 2026
Partial Evidence Bench: Benchmarking Authorization-Limited Evidence in Agentic Systems
arXiv · 6 May 2026
GAGPO: Generalized Advantage Grouped Policy Optimization
arXiv · 13 May 2026
SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution
arXiv · 18 May 2026
How to cite this record
ethics.ai (9 May 2026), “Internal vs. External: Comparing Deliberation and Evolution for Multi-Agent Constitutional Design,” evidence record 4653, https://ethics.ai/record/4653 (originally published by arXiv).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.