Value-Action Alignment in Large Language Models under Privacy-Prosocial Conflict

2026-01-07

arXiv.org

AI Summary

This study addresses the tension between privacy concerns and prosocial motivations in large language models (LLMs) when making decisions involving personal data sharing, where the alignment between expressed values and actual behavior remains unclear. The authors propose a contextually coherent conversational evaluation protocol that sequentially assesses an LLM’s privacy attitudes, prosociality, and willingness to share data. They introduce the Value–Action Alignment Rate (VAAR), a metric quantifying behavioral–attitudinal consistency under competing values by measuring directional alignment relative to human references. Integrating standardized questionnaires, context-preserving dialogue mechanisms, and multi-group structural equation modeling (MGSEM), the research identifies a stable spectrum of privacy–prosociality–sharing acceptance traits across major LLMs and reveals significant inter-model differences in value–action alignment.

Technology Category

• Humans and AI: Learning Human Values and Preferences
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Security and Privacy: Large-scale security measurements

Abstract

Large language models (LLMs) are increasingly used to simulate decision-making tasks involving personal data sharing, where privacy concerns and prosocial motivations can push choices in opposite directions. Existing evaluations often measure privacy-related attitudes or sharing intentions in isolation, which makes it difficult to determine whether a model's expressed values jointly predict its downstream data-sharing actions as in real human behaviors. We introduce a context-based assessment protocol that sequentially administers standardized questionnaires for privacy attitudes, prosocialness, and acceptance of data sharing within a bounded, history-carrying session. To evaluate value-action alignments under competing attitudes, we use multi-group structural equation modeling (MGSEM) to identify relations from privacy concerns and prosocialness to data sharing. We propose Value-Action Alignment Rate (VAAR), a human-referenced directional agreement metric that aggregates path-level evidence for expected signs. Across multiple LLMs, we observe stable but model-specific Privacy-PSA-AoDS profiles, and substantial heterogeneity in value-action alignment.

Problem

• value-action alignment
• privacy
• prosocial behavior
• large language models
• data sharing

Innovation

• Value-Action Alignment
• Privacy-Prosocial Conflict
• Structural Equation Modeling
• Large Language Models
• VAAR