I Can't Believe It's Corrupt: Evaluating Corruption in Multi-Agent Governance Systems

TL;DR

This paper empirically evaluates how different governance structures influence corruption in multi-agent AI systems, emphasizing the importance of pre-deployment integrity checks and institutional design for safe AI delegation.

Contribution

It provides systematic empirical evidence that governance structure impacts corruption outcomes more than model identity, highlighting the need for pre-deployment safeguards and stress testing.

Findings

Governance structure significantly affects corruption outcomes.

Lightweight safeguards can reduce but not eliminate risks.

Institutional design is crucial before deploying AI agents with authority.

Abstract

Large language models are increasingly proposed as autonomous agents for high-stakes public workflows, yet we lack systematic evidence about whether they would follow institutional rules when granted authority. We present evidence that integrity in institutional AI should be treated as a pre-deployment requirement rather than a post-deployment assumption. We evaluate multi-agent governance simulations in which agents occupy formal governmental roles under different authority structures, and we score rule-breaking and abuse outcomes with an independent rubric-based judge across 28,112 transcript segments. While we advance this position, the core contribution is empirical: among models operating below saturation, governance structure is a stronger driver of corruption-related outcomes than model identity, with large differences across regimes and model--governance pairings. Lightweight safeguards can reduce risk in some settings but do not consistently prevent severe failures. These results imply that institutional design is a precondition for safe delegation: before real authority is assigned to LLM agents, systems should undergo stress testing under governance-like constraints with enforceable rules, auditable logs, and human oversight on high-impact actions.