Peer Oversight in Collective Decision Making

Sarah Mohsen; Pavel Naumov

cs.GTAug 28, 2026

Peer Oversight in Collective Decision Making

Authors: Sarah Mohsen, Pavel Naumov

Organizations: School of Electronics and Computer Science University of Southampton, United Kingdom

Abstract

This article introduces peer $k$ -oversight, a property of sequential collective decision mechanisms requiring at least $k$ agents to be responsible for every harmful outcome. It is shown that whenever $k$ -oversight can be achieved by redistributing control over the decisions in a mechanism, it can be achieved using just $k$ agents. A polynomial-time algorithm is also presented that determines whether such a redistribution exists and, when it does, constructs one. These results establish peer oversight as a tractable design principle for multiagent decision-making systems.

Figures & tables

Panel of Figure 1 — Figure 1: Three factories example.

Panel of Figure 2 — Figure 2: Three factories example.

Panel of Figure 3 — Figure 3: Six factories example.

Panel of Figure 4 — Figure 4: Two examples of ranking functions for the mechanism in Figure 1(b) .

Figure 5: Irrecoverability index $\iota(u)$ .

Shared Decision-Making

See all 11 papers

Calibrating Conservatism for Scalable Oversight
May 27, 2026William Overman, Mohsen BayatiScalable Oversight AI Agent Safety
Stanford Graduate School of Business, Stanford University, Stanford, CA, USA.
Delegating Authorization to Misaligned Agents: Coalitional Alignment and Safe Control
Sep 14, 2026Natalie Collina, Surbhi Goel, Aaron Roth +1AI Agent Safety Runtime Enforcement for AI Agents
Beyond Byzantine: An Organizational Consensus Algorithm for Self-Interested Agents Under Information Asymmetry
Jul 31, 2026Jiawei Zhang, Jianbo LiuMulti-Agent Coordination Mechanism Design
Department of Agricultural Economics Purdue University West Lafayette, USA · Basic Technology Group, Intelligent Terminal Business Dept. JD Logistics, JD.com Beijing, China

May 27, 2026J/K move · Enter open · S save

Calibrating Conservatism for Scalable Oversight

William Overman, Mohsen Bayati

Stanford Graduate School of Business, Stanford University, Stanford, CA, USA.

Scalable Oversight AI Agent Safety Human Oversight of AI

Open paper AskPDF arXiv

Agentic AI systems capable of autonomous planning and extended environmental interaction pose a fundamental control problem: how can humans maintain meaningful oversight of systems that may exceed their own capabilities? Existing approaches to scalable oversight rely on complex assumptions, remain largely heuristic, or lack practical methods for sequential settings with statistical guarantees. We introduce Calibrated Collective Oversight (CCO), which aggregates diverse auxiliary scoring functions into a penalty measuring deviation from a conservative baseline. Inspired by Attainable Utility Preservation, CCO enables collective conservatism: actions face a penalty proportional to overseer concern, so high-utility actions are still selected when overseers find them unobjectionable and overridden only when concern accumulates. CCO calibrates this conservatism online using Conformal Decision Theory, ensuring that undesirable outcomes remain below a user-specified target threshold with finite-time bounds and no distributional assumptions. On a modified version of SWE-bench, weaker overseers successfully constrain an adversarially misaligned stronger agent; on MACHIAVELLI, CCO substantially reduces ethical violations while preserving reward. In both settings, empirical violation rates closely match the specified targets, as predicted by the theory.

Abstract

Shared Decision-Making

Explore similar work

Calibrating Conservatism for Scalable Oversight

Delegating Authorization to Misaligned Agents: Coalitional Alignment and Safe Control

Beyond Byzantine: An Organizational Consensus Algorithm for Self-Interested Agents Under Information Asymmetry