PADTHAI-MM: Principles-based Approach for Designing Trustworthy, Human-centered AI using MAST Methodology

Explainability and contextual information significantly impact user trust: The study compared two versions of an AI system—one with explanations and AI context (High-MAST) versus a "black box" version (Low-MAST)—and found that providing users with information about how the AI works and why it makes recommendations substantially increased trust in the system for intelligence reporting tasks.

A systematic framework can guide trustworthy AI design: PADTHAI-MM offers an actionable, principles-based methodology for designing AI systems in high-stakes domains, moving beyond theoretical discussions of trust to provide practical design and evaluation tools that developers and organizations can actually implement.

Stakeholder feedback is essential throughout development: The framework emphasizes an iterative design process guided by input from actual users and stakeholders, suggesting that trustworthy AI cannot be designed in isolation but requires ongoing collaboration with the people who will use these systems.

One-size-fits-all approaches are insufficient: The paper highlights that context-specific needs matter—the framework is designed to be adaptable to different domains and use cases rather than applying generic trust principles uniformly across all AI applications.

Trust in AI systems requires more than just accuracy: By focusing on explainability, transparency, and human-centered design rather than solely on technical performance, the research demonstrates that building trustworthy AI involves addressing how users understand and interact with the system, not just whether it produces correct outputs.