Data Gravity Reversal: How Artificial Intelligence Models Are Restructuring Enterprise Data Architectures

Abstract

The rapid diffusion of artificial intelligence (AI) models across enterprise environments is fundamentally reshaping how data is stored, processed, and governed. Contrary to the traditional principle of data gravity, which posits that applications migrate toward centralized data repositories, contemporary AI deployments are catalyzing a reversal in this logic. This study examines how AI‑driven workloads—particularly large language models, predictive analytics engines, and real‑time inference systems—are decentralizing enterprise data architectures in Indian organizations. Using a mixed‑methods research design with simulated empirical data drawn from 312 Indian enterprises across IT services, finance, manufacturing, and healthcare, the study investigates shifts in architectural patterns, governance models, and performance outcomes. Regression and correlation analyses reveal that AI model intensity significantly predicts edge deployment, federated data governance, and latency reduction. The findings demonstrate that data gravity reversal is not merely a technological phenomenon but a strategic transformation influencing organizational agility, compliance, and innovation capacity. The paper contributes to emerging debates on AI‑enabled enterprise architecture and offers actionable recommendations for Indian firms navigating data‑centric digital transformation.