Examining the Impact of Label Detail and Content Stakes on User Perceptions of AI-Generated Images on Social Media

2025-10-21

AI Summary

This study investigates how the granularity of AI-generated image labels (basic/intermediate/maximal) and content risk level (high/low) interactively influence user engagement, transparency perception, and trust in social media contexts. Method: A within-subjects experimental design manipulated label granularity and content risk as core independent variables to empirically assess their effects. Results: (1) Increasing label granularity significantly enhances transparency perception without reducing user engagement; (2) Low-risk content independently and significantly boosts both engagement and trust; (3) Label optimization and risk mitigation exhibit synergistic potential—applying intermediate-to-maximal labeling in low-risk scenarios balances transparency and user experience. This work provides the first systematic empirical evidence of the interaction between label design and content attributes, offering foundational insights and actionable guidelines for developing trustworthy, usable AI content labeling frameworks.

Technology Category

• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Humans and AI: Intelligent User Interfaces
• Computer Vision: Interpretability, Explainability, and Transparency

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

AI-generated images are increasingly prevalent on social media, raising concerns about trust and authenticity. This study investigates how different levels of label detail (basic, moderate, maximum) and content stakes (high vs. low) influence user engagement with and perceptions of AI-generated images through a within-subjects experimental study with 105 participants. Our findings reveal that increasing label detail enhances user perceptions of label transparency but does not affect user engagement. However, content stakes significantly impact user engagement and perceptions, with users demonstrating higher engagement and trust in low-stakes images. These results suggest that social media platforms can adopt detailed labels to improve transparency without compromising user engagement, offering insights for effective labeling strategies for AI-generated content.

Problem

• Examining how label detail affects AI image transparency
• Investigating content stakes impact on user engagement
• Analyzing user perceptions of AI-generated social media images

Innovation

• Detailed labels enhance transparency without reducing engagement
• Content stakes significantly influence user trust and engagement levels
• Platforms can implement detailed labeling for AI-generated images