Model Cards Revisited: Bridging the Gap Between Theory and Practice for Ethical AI Requirements

2025-07-08

AI Summary

Current AI model cards exhibit a systemic gap in ethical practice, overemphasizing performance and reliability while neglecting core ethical dimensions—including explainability, fairness, and user autonomy. This study systematically analyzes 26 international ethical guidelines, three mainstream documentation frameworks, and representative model cards, applying thematic analysis to identify 43 critical ethical requirements. It proposes a comprehensive ethical taxonomy organized around four overarching themes—Transparency, Fairness, User Empowerment, and Accountability Governance—comprising twelve subthemes. The analysis reveals systematic deficiencies in existing documentation across breadth, depth, and operationalizability of ethical coverage. To address these gaps, we introduce an enhanced model card framework that provides developers with a structured, actionable template for recording and disclosing ethically relevant information. This framework significantly improves the visibility and verifiability of AI systems’ ethical compliance, advancing practical implementation of responsible AI development.

Technology Category

• Philosophy and Ethics of AI: Accountability, Interpretability & Explainability
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy
• Machine Learning: Ethics, Bias, and Fairness

Application Category

• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization
• Responsible Web: Fairness, accountability, transparency and ethics of web technologies
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Model cards are the primary documentation framework for developers of artificial intelligence (AI) models to communicate critical information to their users. Those users are often developers themselves looking for relevant documentation to ensure that their AI systems comply with the ethical requirements of existing laws, guidelines, and standards. Recent studies indicate inadequate model documentation practices, suggesting a gap between AI requirements and current practices in model documentation. To understand this gap and provide actionable guidance to bridge it, we conducted a thematic analysis of 26 guidelines on ethics and AI, three AI documentation frameworks, three quantitative studies of model cards, and ten actual model cards. We identified a total of 43 ethical requirements relevant to model documentation and organized them into a taxonomy featuring four themes and twelve sub-themes representing ethical principles. Our findings indicate that model developers predominantly emphasize model capabilities and reliability in the documentation while overlooking other ethical aspects, such as explainability, user autonomy, and fairness. This underscores the need for enhanced support in documenting ethical AI considerations. Our taxonomy serves as a foundation for a revised model card framework that holistically addresses ethical AI requirements.

Problem

• Bridging gap between AI ethical requirements and documentation practices
• Improving model cards to include overlooked ethical aspects
• Providing taxonomy for ethical AI documentation in model cards

Innovation

• Thematic analysis of ethics guidelines and model cards
• Taxonomy of 43 ethical requirements for AI documentation
• Revised model card framework addressing ethical principles