Data Gravity vs. Model Agility: The New Tension Shaping the Future of Automation and AI: A Systematic Review

Abstract

Background: The revolutionary advancement of AI in automation as proved with an exponential development in AI has led to a new paradigm shift in computing. As the pull of data gravity (e.g., massive datasets pooled centrally) collides with the model agility of decentralized, edge-based learning, an important tension emerges. This article integrates current academic and operational perspectives on this budding dichotomy and its implications for the future of AI-based systems. Objective: In this literature review, we take a closer look at the history of, and trends in, the data gravity versus model agility discussion, from a conceptual, empirical and technological perspective. It seeks to examine how friction characterizes automation, agility, infrastructure design, ethical governance, and the deployment of AI. Methods: Methods We performed a systematic review and synthesis of >60 (policy white papers and operational concept reports, before March 2025) peer-reviewed articles and operational concept reports. What key thematic lenses are relevant for viewing algorithmic systems in the context of datafication process? These might include data localization, edge computing, federated learning, smart data strategies as well as institutional AI design? Results: The review extracts three dominant narratives of transformation: (1) the near-singularity of data-driven infrastructures, (2) the ascendancy of edgeletigence as an unseen frontier in automation, and (3) the death of the big data towards the rise of could-aware, low-latency smart data systems. Study results, levels of evidence, conceptual models, and guidelines are summarized in tables. Conclusion: A clear trend is emerging around hybridized AI system that combines data locality and model distribution. Systematically integrating ethical oversight, agile architectures and human-machine collaboration in strategic terms, is an increasingly pressing research and policy concern.