Methodological, Data-Related, and Governance Challenges

Abstract

This chapter critically examines the methodological, data-related, and governance challenges that constrain the effective and responsible implementation of Artificial Intelligence in Sports Science. It positions these limitations as a necessary counterbalance to prevailing technological optimism and situates them within both practical modelling workflows and broader epistemological considerations. Building on previous chapters, the analysis emphasizes the importance of validation, robustness testing, and sensitivity analysis as essential components of reliable AI systems. The chapter identifies key barriers to implementation, including economic constraints, limited access to qualified personnel, and institutional resistance-driven by cultural skepticism toward algorithmic decision-making. It further explores the tension between underutilization and overreliance on AI, highlighting risks associated with automation bias and the erosion of human expertise. From a governance perspective, the chapter examines the need for standardization, certification, and auditability, referencing frameworks such as the EU Artificial Intelligence Act and NIST guidelines. It also addresses issues of accessibility and equity, advocating for the democratization of AI technologies through open-access tools and federated learning approaches. The chapter concludes by emphasizing sustainability, stakeholder trust, and the integration of ethical, educational, and regulatory mechanisms as prerequisites for long-term AI adoption in Sports Science.