Who Owns That? Evaluating Ownership Intuitions in Large Language Models

2026-09-30

AI Summary

This study investigates whether the ownership intuitions of large language models (LLMs) under competing claims align with those of humans. We introduce COAT, a benchmark that systematically evaluates the allocation decisions and contextual sensitivity of 24 LLM configurations against 108 human participants through behavioral experiments and statistical comparative analyses. Results indicate that although LLMs broadly approximate human judgment trends, their decisions are notably more homogeneous and biased toward equal division. Furthermore, the models’ responses to contextual factors such as item value and public recognition diverge significantly from human behavior, lacking the diversity, heterogeneity, and context-dependence inherent in human reasoning. These findings reveal critical limitations in LLMs’ capacity for complex social norm reasoning and offer new perspectives for model alignment research.

Abstract

Ownership establishes rights over the use, control, and transfer of objects. Understanding these relations is essential for AI systems to interact appropriately with people and their resources. Yet how large language models (LLMs) attribute ownership under competing claims remains unclear. We introduce the Competing Ownership Attribution Task (COAT), comprising 42 scenarios, and compare ownership allocations from 24 LLM configurations with those of 108 human participants. Overall, human-model similarity is close to human-human similarity, but models show greater homogeneity in their ownership judgments. Within individual answers, models also divide ownership more evenly among claimants than humans do. Pooling responses across model configurations reveals more scenarios with a shared judgment and fewer with distinct viewpoint groups than in humans. When humans form distinct groups, models may converge on one viewpoint or between competing viewpoints. Further comparisons reveal different contextual sensitivities. As material value increases across scenarios, allocations to creators decline less sharply in models than in humans. Across scenarios differing in public recognition of later holders as owners, allocations to these holders increase in models but decrease slightly in humans. Together, these findings suggest that the evaluated LLM responses do not fully capture the diversity of participants' ownership judgments or how those judgments vary across situations. Developing socially capable AI therefore requires moving beyond overall similarity to capture the diversity and context dependence of human judgments.

Problem

• ownership attribution
• large language models
• human-AI alignment
• social reasoning
• context dependence

Innovation

• Ownership Attribution
• Large Language Models
• COAT Benchmark
• Human-AI Alignment
• Context Sensitivity