"I Don't Think RAI Applies to My Model'' -- Engaging Non-champions with Sticky Stories for Responsible AI Work

2025-09-26

AI Summary

Responsible AI (RAI) advocates often perceive ethical tools as redundant bureaucratic overhead, exhibiting low sensitivity to AI-related risks. Method: Drawing on ethnographic fieldwork, this study derives RAI design principles tailored specifically for such practitioners and introduces the novel “sticky story” intervention paradigm—narratives that are concrete, severe, unexpected, and multi-faceted in their depiction of AI harms. To enable scalable, verifiable story generation, we propose a human-AI co-creation framework integrating large language model (LLM) automation with expert human evaluation. Contribution: In an experimental study with 29 practitioners, sticky stories significantly increased risk identification latency, broadened the range of harm types recognized, and deepened reflective engagement. This constitutes the first empirical validation that narrative-driven interventions effectively activate ethical awareness among non-advocates—a critical advance toward operationalizing RAI through behaviorally grounded, generalizable interventions.

Technology Category

• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Humans and AI: Intelligent User Interfaces

Application Category

• Responsible Web: Fairness, accountability, transparency and ethics of web technologies
• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization

Abstract

Responsible AI (RAI) tools -- checklists, templates, and governance processes -- often engage RAI champions, individuals intrinsically motivated to advocate ethical practices, but fail to reach non-champions, who frequently dismiss them as bureaucratic tasks. To explore this gap, we shadowed meetings and interviewed data scientists at an organization, finding that practitioners perceived RAI as irrelevant to their work. Building on these insights and theoretical foundations, we derived design principles for engaging non-champions, and introduced sticky stories -- narratives of unexpected ML harms designed to be concrete, severe, surprising, diverse, and relevant, unlike widely circulated media to which practitioners are desensitized. Using a compound AI system, we generated and evaluated sticky stories through human and LLM assessments at scale, confirming they embodied the intended qualities. In a study with 29 practitioners, we found that, compared to regular stories, sticky stories significantly increased time spent on harm identification, broadened the range of harms recognized, and fostered deeper reflection.

Problem

• Engaging non-champions with Responsible AI tools
• Addressing practitioner perception of RAI irrelevance
• Improving harm identification through sticky stories

Innovation

• Using sticky stories to engage non-champions
• Generating narratives via compound AI system
• Evaluating stories through human and LLM assessments