Prompt framing governs LLM default following in collective-action

2026-10-02

AI Summary

This study investigates whether large language models (LLMs) exhibit blind conformity to preset defaults in collective action scenarios and how prompt framing shapes their decision-making behavior. Leveraging common-pool resource and public goods games, the authors employ single-shot simulations, probability mass shift measurements, and multivariate controlled experiments to quantify the biasing effects of default options on model choices across varying phrasings and action space granularities. The findings reveal that default compliance is highly contingent upon interface wording and action space structure, with permissive formulations significantly mitigating blind adherence. Furthermore, this work demonstrates that choice architecture profoundly modulates behavioral variance in LLMs, underscoring the necessity of rigorously controlling default option design when evaluating agent-based systems.

Abstract

Large language models are increasingly deployed as agents that make or recommend decisions on behalf of users, often operating through interfaces that pre-fill suggested values or default options. Whether models treat such defaults as merely informational or as suggestions that systematically alter their choices remains unclear. We study default deference in two one-shot social dilemmas: a common-pool resource (CPR) extraction game and a threshold public-good (TPG) contribution game. We measure how defaults shift each model's choice distribution relative to its no-default baseline, across default values, wordings, and action-space granularities. We find that pre-filled defaults pull probability mass on the default value in both games, but the magnitude depends strongly on wording: the same model can show high pull under one formulation and near-zero pull under another. Permission-style wording reduces default pull in both games, more strongly in CPR than in TPG. Default pull is weaker in coarse action spaces, and conflict defaults attract more mass than agreement ones. These results indicate that default deference depends on the model, the wording of the interface, and the structure of available choices. For agentic systems, evaluating model behaviour without controlling the surrounding choice architecture can miss an important source of behavioural variation.

Problem

• Large Language Models
• Default Deference
• Collective Action
• Social Dilemmas
• Prompt Framing

Innovation

• Large Language Models
• Default Deference
• Prompt Framing
• Collective Action
• Choice Architecture